Old Dogs, Old Tricks. Dealing with TLT Spam

By KineticOperator, in X-Wing

As I've no clue what "Rebel Convoy" is I guess I'm going to have to look into another pair of B-Wings. ;) Will it still work if I drop the Z-95 to buff the Bs?

As I've no clue what "Rebel Convoy" is I guess I'm going to have to look into another pair of B-Wings. ;) Will it still work if I drop the Z-95 to buff the Bs?

Again, it's IN THE OP. I put some effort into that post, because I try to help new folks and it doesn't do any good if it's nothing but jargon. Please take a moment to actually read through it. I hope that doesn't come across as rude, but this is the second time you have seemingly ignored the actual content of the thread/OP.

As for dropping the Z to buff the B's; It can work. However, by doing so you dramatically reduce your advantage in overall health, give up an advantage in attack dice, and lose a potential blocker with which to deny your opponent space. To give up all of that you would need to put some upgrades on those B-Wings that offer very significant advantages if you wish to remain as effective against TLTs. On the other hand, maybe you will decide that increased effectiveness against some other list (depending on what you are upgrading against) will be worth the loss of effectiveness vs. TLTs.

I suppose the best advice is that when upgrading, upgrade with a purpose. Most people simply put stuff on that would make the ship better, without considering what it makes the ship better AGAINST. If your upgrade only makes you even better against an uncommon list you are already 80/20 against it is almost certainly not worth the points. Put those points into improving your odds against lists you expect to see, or that you are vulnerable against (get to at least 40/60 with 50/50 as your goal) rather than simply piling on. "Win More" upgrades are seldom worth the points.

Edited by KineticOperator

@SanguinaryDan: there's also this thread, which I put together a few weeks ago to address questions like yours. My advice isn't as specific as KO's, but it may be helpful as you figure out what you'd like to run.

Yes it is. With something called ORS. But since I appear to have comprehension issues with your posts I won't ask what the hell that is.

I'll give the 4BZ list a shot. If I can make it work in random games I'll try it in the tourney.

In related news, fellow NOVA podcast co-host Kris just got back from a 25 person tournament. There was only one 4 TLT list, but it won (4 Y Scum). It beat a 3A Procket list twice. Not sure what else it faced.

Yes it is. With something called ORS. But since I appear to have comprehension issues with your posts I won't ask what the hell that is.

I'll give the 4BZ list a shot. If I can make it work in random games I'll try it in the tourney.

I didn't mean any of that as a dig, at all, I was just getting the impression you hadn't actually read the OP. Sorry about the jargon, it crept in regardless. :-)

ORS is Outer Rim Smuggler, the generic YT-1300. Ion Warthogs are BTL-A4 Y-Wings with Ion turrets.

3A Prockets would be all of the challenges of the Royal Flush, without barrel roll or a consistent 3 attack dice. Without something more to help up the damage output like the upcoming Juke, it would be one heck of a challenge to beat 4 TLTs with. Heck, even with Juke it will be an uphill battle, A-Wings just don't throw the dice needed to hammer down 32 points of shields and hull.

Edited by KineticOperator

I would add, also, that in the KO's TC Aces match of 3 Interceptor Aces vs TLTs, he had extraordinarily good luck especially in the first couple of combat rounds. I don't have the exact numbers handy, but if the luck went equivalently the other way in favor of TLT, then at a minimum Carnor Jax would have been dead in the 2nd combat round with all 4 Y-wings still on the table. It can be enticing to look at a single game as an example of how it can be done, but you have to normalize the results to the round-by-round luck of the dice rolls, and extrapolate what a more "normal" distribution of results would be.

What I think would be a hilarious anti-TLT list would be 4 shuttles and a bomber. 46 hull points for TLTs to chew through!

What about Scum? Apart from Rebel lists they can pull off as well (8 z-95s) and Brobots, what would be a good Scum list to take on TLTs?

I said it before if TLT are a problem learn tofly interceptors with autothrusters and teach thoes TLT players a leason

I have two non-standard-format tournaments coming up in the next several weeks and expect a meta wherein at least 50% of the lists played will include some number of TLT Y-Wings. I've already found the OP to be immensely helpful, but have a few questions specific to these upcoming tournaments.

The first one is a two-person team tournament with each teammate fielding her/his her own 100pt squad. The squads must be the same faction and will be considered friendly to one another. Duplication of unique pilots or upgrades is, unsurprisingly, not allowed. It's important to note (because this seemed to be missed frequently when I posted about this previously) that this is NOT a 200pt squad, but two separate 100pt squads. In reference to the OP, my question is this: is it recommended to double-down on one of the squads generic pilots, or would my teammate and I be better served to combine two different lists of the same faction from the OP? I'm leaning toward the BBBBZ and/or Rebel Convoy lists.

The subsequent tournament is a 120pt single-player tournament. As above, I'm leaning toward the BBBBZ list, and I have a couple of closely related questions about adapting it to a 120pt format: 1) is the Z important as anything other than filler, or would we prefer another B if we had the points? I notice 5 Daggers comes in at exactly 120pts and the PS bid seems like it works for me against TLT spam, but would I do better with 5x Blue + FCS?

Any insight, especially from OP, is greatly appreciated.

Yes, but.....

Is there a currently legal Rebel build that can cope with quad TLT and still handle what ever else a tournament might throw at you?

I'll posit - that at best, a balanced, and diverse game will not have any single list that can "Handle" whatever a tournament might throw at you. If by "Handle" we mean have a good matchup (win more times than it loses based on the list alone) - and that if the balance becomes truly rich, balanced and diverse - that there will be an equal number of good and bad matchups out there - and that player skill will surface out of the noise of the dice - and that bad-matchups can be won beyond just "luck of the dice".

While I doubt X-wing is balanced on a pinhead, that it's reasonably close enough at the moment that usually player skill can overcome luck, with a wide variety of builds. If you're looking for something that has a wide variety of matchups - you're really looking for "the best list" which I hope - doesn't exist, because then there's not much reason to fly anything else (competitively)

I would add, also, that in the KO's TC Aces match of 3 Interceptor Aces vs TLTs, he had extraordinarily good luck especially in the first couple of combat rounds. I don't have the exact numbers handy, but if the luck went equivalently the other way in favor of TLT, then at a minimum Carnor Jax would have been dead in the 2nd combat round with all 4 Y-wings still on the table. It can be enticing to look at a single game as an example of how it can be done, but you have to normalize the results to the round-by-round luck of the dice rolls, and extrapolate what a more "normal" distribution of results would be.

Well...

Looking at the dice, my Interceptors rolled 2 "extra" blank greens in the first round of shooting, vs. 3 "extra" blank hits rolled by the Y-Wings. Even if you give another hit to Matt, it would still be a total of 13 Interceptor evades vs. 12 Y-Wing hits. On the other side, my reds vs. his greens were exactly average. The fact that Jax took 2 damage in this exchange does not suggest the Interceptors dice were particularly good.

In the second round, Interceptors rolled .5 "extra" blank greens vs. 1.5 "extra" blank Reds from the Y-Wings, and on the other side Interceptors rolled .62 "fewer" blank reds vs. his .5 "extra" blank greens.

In other words, it wasn't actually the dice rolls that altered that exchange, it was the presence of focus tokens and Autothrusters. Putting Interceptors out of arc and/or at range 3 of every shot coming their way makes it very difficult for TLTs to get hits through. The skew in results over the first couple rounds was due almost entirely to the fact that at least half the TLT shots were completely unmodified, where Interceptor defenses were universally boosted by Autothrusters and sometimes with Focus/Evade as well.

By the end of the game, Interceptor dice rolls were dead average, offensively and defensively. Y-Wing rolls had not been as good, at 6.5 "extra" blanks on attacks and 3.25 "extra" blanks on greens. But considering the Y-Wings had 110 red dice rolled and 22 green dice rolled, this is only a variance of 5.9% offensively and 14.7% defensively, well within expected ranges.

Finally, the game didn't end close. Jax was down to a single Hull, but Turr still had 2 and the Y-Wings had yet to even take a shot at Soontir. Assuming you reversed the dice, it would still have ended with a solid Interceptor win, and this is with well handled Y-Wings in the hands of a world class player.

Matt didn't make big mistakes (or little ones, frankly) that I exploited. He was limited by the low skill ceiling of Y-Wings. There were only a very limited number of options for him when maneuvering, and Interceptors are well equipped to outmaneuver predictable ships. Had a player of his caliber been flying something with more options, it would have been much harder for me to anticipate his moves and maneuver in a way that allowed Autothrusters and Jax's ability to "skew" the dice in my favor. As long as the Interceptor player ensures that he always has the benefit of Autothrusters, he can expect games to go well a very large percentage of the time (I put it at about 80/20 in favor of the Interceptors, barring mistakes).

The last point that I don't see discussed often is that TLT lose the ability to benefit from outlier attacks. Averages are interesting, but games turn on the "odd" rolls. For example, one player rolls a natural 2 hits + crit and the other blanks his defense. One of the greatest weaknesses of Interceptors is their vulnerability to a single outlier roll. Facing TLTs, you no longer have to worry about these outliers hurting you but you still have the opportunity to inflict them. It's a very big deal, especially when you start making a lot of attacks, because while the odds of an outlier occurring in any single roll are low, in the aggregate the odds are nearly certain. In this game, there was an early "hit, crit, crit" roll vs. no defense at Jax that would have killed him outright, except it was a TLT shooting so Jax took 1 generic damage and survived unimpeded. Later, there was a "hit, hit, hit, crit" roll vs. "blank" on a 5 health Y-Wing that resulted in a Direct Hit and instant death. Both of these are potential major turning points, and the Jax hit occurred much earlier, but due to the nature of TLT damage the "game of outliers" fell squarely on the side of the Interceptors.

TL:DR, The Interceptor vs. TLTs game went pretty much exactly as expected. Though Jax should have died at some point, it wouldn't have altered the final outcome. Because reasons. :-)

Edited by KineticOperator

Looking at the dice, my Interceptors rolled 2 "extra" blank greens in the first round of shooting, vs. 3 "extra" blank hits rolled by the Y-Wings. Even if you give another hit to Matt, it would still be a total of 13 Interceptor evades vs. 12 Y-Wing hits. On the other side, my reds vs. his greens were exactly average. The fact that Jax took 2 damage in this exchange does not suggest the Interceptors dice were particularly good.

In the second round, Interceptors rolled .5 "extra" blank greens vs. 1.5 "extra" blank Reds from the Y-Wings, and on the other side Interceptors rolled .62 "fewer" blank reds vs. his .5 "extra" blank greens.

In other words, it wasn't actually the dice rolls that altered that exchange, it was the presence of focus tokens and Autothrusters.

It's not just the number of hits / evades, it's the total probability density function of how much damage a given ship should have taken over the course of the entire turn (or multiple turns). Veteran playtesting is your domain of expertise, and technical analysis is mine, so lets jump right in. :)

Round 1

The attacks onto the Y-wings were exactly average at 3 damage vs 3.03 expected damage, but Jax got a little lucky. The average expected damage onto Jax in round 1 was 2.7 vs an actual of 2. Specifically, the exact PDF is:

0.0258 0.1271 0.2613 0.2918 0.2939

So Jax had a 29.39% chance of getting outright killed in the first round, and had a 58.57% chance of taking 3+ damage in the first round. There was a 41% chance that you take 2 damage or less in the first round.

Incidentally, Matt should have rolled 3 dice on his defense vs Soontir Fel (range 3 through a rock, but he only rolled 2 green dice), but it didn't affect anything.

Round 2

Matt helped you out big time in round 2 by taking target lock actions on Fel instead of target locking Jax, or even simply taking the focus action, which would have helped him kill off Jax. This needs to be considered if you want to analyze this match properly. If you fly better than your opponent, or equivalently if your opponent makes mistakes, then of course you should be expected to win. To analyze matchups as unbiased as possible, you need to normalize player skill. If Matt instead takes focus with his 3 Y-wings in round 3, and the Y-wing who ends up eating munitions failure spends focus on defense, then the odds of outright killing Jax by round 2 are:

  • with munitions failure: 70.8% (3.56 average damage, capping damage at 4)
  • without munitions failure: 78.3% (3.68 average damage, capping damage at 4)

Conversely, the chance of Jax having 2+ hull remaining (as he did in the actual match) after the 2nd round (if Matt focuses) is:

  • with munitions failure: 11.4%
  • without munitions failure: 7.8%

If Matt focuses round 2, then Jax is twice as likely to to be dead on the 2nd round than he is to have been alive with 2+ hull alive. If the luck were to have merely been "fair", then Jax would have been dead in the 2nd round. If the luck through round 2 had instead swung in favor of Matt to the same degree that it swing to your favor, then Jax absolutely would have been dead, and probably none of the Y-wings would have taken any hull damage, and might have even still had some shields remaining on Yasmina.

Analyzing the game as the actions played out, where he only focused on one Y-wing that had a shot on Jax, then the chances of killing Jax through round 2 drop slightly to:

  • with munitions failure: 62.9% (3.42 average damage, capping damage at 4)
  • without munitions failure: 71.8% (3.57 average damage, capping damage at 4)

Likewise, the chance of Jax having 2+ hull remaining (as he did in the actual match) after the 2nd round (if Matt target locks Fel) is:

  • with munitions failure: 16.0%
  • without munitions failure: 11.2%

All of this analysis ignores the probability that Jax gets killed before the last shot in round 2, and the Y-wings can start taking shots at either Turr or Fel, which would marginally improve the Y-wings' outlook.

Round 4

If Jax dies in round 3, and the Y didn't eat munitions failure, then actionless Turr gets hit for an average of 1.38 damage next round from three TLTs (assuming Yvonne K-turns to get a shot instead of hard left 2), with a 15% chance of dying outright and a 22% chance of escaping that round unscathed. Instead, Jax was still alive at 1 hull, Turr was untouched, and Jax was still doing damage (this round and the next) to the Y-wings.

Round 5

Fel does 5 damage to a Y-wing to take it off the table, Jax does 2 more damage to a Y-wing, and then Jax (who should have been dead 2 rounds ago) rolls 3 natural evades to stay alive Yet Again. The game was already skewing heavily in favor of KO because of dice, but this put it over the edge. I don't think I need to run the odds to make it obvious. ;)

By the end of the game, Interceptor dice rolls were dead average, offensively and defensively. Y-Wing rolls had not been as good, at 6.5 "extra" blanks on attacks and 3.25 "extra" blanks on greens. But considering the Y-Wings had 110 red dice rolled and 22 green dice rolled, this is only a variance of 5.9% offensively and 14.7% defensively, well within expected ranges.

You can't simply look at rolls at the end of the game, as rolls in the beginning have a far greater impact at the final state than rolls near the end of the game. Real-world examples: the time value of money, compound interest, etc. Once you get the snowball effect going, it can be almost impossible to overcome later. To reverse initial bad luck you will need good luck that is even more of an outlier than your early bad luck. To get a true appraisal of the effect of luck, you need to take an integral of the luck per round multiplied by a time based weighting function. Looking at Lady Luck's final singular number is unfortunately a very poor way of judging luck.

Finally, the game didn't end close. Jax was down to a single Hull, but Turr still had 2 and the Y-Wings had yet to even take a shot at Soontir. Assuming you reversed the dice, it would still have ended with a solid Interceptor win, and this is with well handled Y-Wings in the hands of a world class player.

Matt didn't make big mistakes (or little ones, frankly) that I exploited.

Again, it wasn't close because Matt did in fact make a mistake in round 3 (TL Fel vs Jax, or at least not focusing) and you also got lucky to have Jax survive, and get a munitions failure on a Y. Once you get to this critical point, things snowball quickly. 4 Y-wings with shields vs 2 Interceptors is a completely different game than how it played out.

The last point that I don't see discussed often is that TLT lose the ability to benefit from outlier attacks. Averages are interesting, but games turn on the "odd" rolls.

Right, and this is implicitly calculated in the above analysis through the 3rd round.

TL:DR, The Interceptor vs. TLTs game went pretty much exactly as expected. Though Jax should have died at some point, it wouldn't have altered the final outcome. Because reasons. :-)

So here is where I am leading to with my final point. The game did indeed go exactly as KO expected. However, a more rigorous analysis of this match demonstrates that if the luck were "fair" and Matt didn't make a key mistake early on, then Aaron probably should have lost. If Matt played well and the luck swung in Matt's favor instead of Aaron's, then Aaron would have lost decisively. Aaron's failure to recognize this is indicative of lacking the technical capability to quantitatively fact check the match results, and also, in my opinion, indicative of confirmation bias.

I would like to go out of my way here to pause, and say that Aaron having confirmation bias is no fault of his own. Even for a veteran player it is nearly impossible to properly analyze a match without the right technical tools. Aaron is a great player and I am sure that his contributions to the game as a playtester are significant. There is a reason why he has made the cut at Worlds for two years in a row, no small feat indeed. :)

Aaron's domain of expertise is being a veteran playtester, and my domain of expertise is in technical analysis. In this particular instance he has ventured into a domain that is not his expertise. So here is the real point:

Proper game balance and playtesting cannot be achieved without using the correct technical tools, and through seven waves of releases there is no evidence that FFG or its playtesters have sufficient technical capability to use let alone invent such tools.

This is not a problem with the playtesters or even the game designers, but rather a high level business decision to not hire personal with this technical capability. Advanced mathematics is not traditionally a requirement for a game designer at such a company, and a job candidate with such qualifications would command a salary well outside FFG's pay rates for these positions. In the meantime, we are all still buying their plastic toys, so their strategy seems to be working well enough.

What it does mean, however, is that when a veteran player such as Aaron publicly makes a post essentially stating "we are all veteran playtesters, XYZ is not broken, trust us", I remain wholly unconvinced despite Aaron's best intentions, significant street credibility, and documented expertise. As it relates to TLT in particular, I am still reserving judgment until I finish some further analysis, although preliminary analysis points to TLT being very good, but probably not quite as meta-warping as the pre-nerf ACD Whisper. For better or for worse, I have no current ETA for finishing said analysis, although even if it were to be done now I would most likely withhold it until after Worlds.

TL; DR: Despite the help of many excellent players such as Aaron in playtesting, game balance design will continue to have a reasonably large variance unless FFG can employ some properly executed advanced mathematics in their design and playtesting process.

Edited by MajorJuggler

I see your points Juggler, but I disagree that this is a case of confirmation bias. You generally look at things in a straight line, jousting formula, but the game is NOT a straight line, dice rolling joust (at least not entirely). The idea that Matt made "mistakes" is something you are adjudicating from the point of view of knowing what did happen, as opposed to what might have happened.

Matt took TLs, because he self bumped his Y-Wing in order to keep Jax where he was and deny the more likely 2 bank by Jax. By doing so, he left himself open if I had instead done a 2 turn. Had he guessed wrong and Jax had revealed a 2 turn, Jax would easily have been able to deny all 3 of his Y-Wings the use of their action, have focus and evade to survive a single actionless attack, as well as sit in a spot where none of the lead 3 could shoot at Jax himself. Knowing that his planned movement + a 2 turn on my part would have utterly wrecked him he made the correct choice and took TLs on a ship he knew he would need accuracy for, knowing that no matter what those TLs could be used at some point in the game. TLs were the correct choice, because they could be left for a later turn when stacked actions would be necessary to penetrate Fel's Stealth device.

Also, on turn 1 Jax had a 29% chance of dying, and a 41% chance of having exactly what happened, happen. This is something I was aware of when I boosted and barrel rolled him out of arc, and knowing I was much more likely to come through it than not is why I made the decision I did.

The next round, IF I had only had 1 hull I would have made a 3 turn rather than clear stress. A stress clear with a high probability of bump is a play for a subsequent round, not this one. A 3 turn would have put Jax dead on his formation no matter where he went, but with a high likelihood of ending up in front of a Y-Wing at close range. With no actions and no autothrusters, that move would have been a sacrifice play to put more damage on his wounded Y-Wing, potentially deny him focus, and draw his fire away from my other 2 ships. Further, had I planned for the sacrifice play I would have had Soontir turn inside the rock, rather than outside, knowing he would be obligated to attempt to finish off Jax. 4 additional dice into his Y would have had him dead, not just munitions fail. Because I had 2 hull, I played for subsequent turns, gambling that 6 shots from TLT would leave Jax intact.

In the real game, a munitions failure made it very unlikely that Jax would go down because those 6 shots suddenly became 4. That is another point lost in the averages. Outlier results of some sort were "guaranteed" by that point (as much as anything based in probability can be). In fact, several had gone by already but they had favored the Y-Wings and TLTs cannot capitalize. This isn't "luck", this is risk management. The odds of a thing happening are one piece of information, but the consequences inherent in that thing happening are at least as large of a concern.

I am not intending to deny the math, though I do contend that saying dice favored me "heavily" in the first two turns when in reality there was only a single point of damage swing between expected incoming damage and actual incoming damage is misleading. Saying that "if the munitions failure had not happened" is also misleading, because lopsided vulnerability to outliers/crits is one of the greatest disadvantages TLT lists share. Lastly, saying "if Matt had not taken TL on Fel" ignores the fact that TLT lists are terrible at reacting. Matt was making solid choices based on limited options and limited information. My moves might have looked "better" by comparison, but in reality it was no more than the fact that my list was extremely reactive and I was making my choices with much more complete information.

The point I am making is that as useful as raw statistical analysis is, it does not reflect the adaptive nature of the game. The most mathematically inefficient lists also happen to be the lists most capable of reacting to changing circumstances, which throws the numbers off dramatically. The list I was using is a very extreme example of this, and as such is able to capitalize tremendously. On the other hand, in terms of raw efficiency it is NOT a good match, which is why it is at the very top of my difficulty list.

I use your analysis extensively, it is critical to know what to expect. Not only in terms of averages, but in the probability of disastrous/beneficial outliers as well. My play has improved tremendously because I have been able to use your numbers, rather than making incomplete calculations on my own (because I have neither the time nor the interest in the depth of analysis you perform). But being effective with ostensibly "bad" lists is a passion of mine, and it isn't consistently good luck that makes them work. I much prefer capitalizing on non-math advantages (couldn't think of a better way to describe it) like base size, maneuverability, etc., than raw efficiency. It isn't better, it's just what I prefer.

Of course, the reason I have so much trouble winning all the way through tournaments is because by relying on "other than efficiency" factors I leave myself open to losing on a single bad decision, or a single bad die roll. That's OK, because I enjoy it, but if someone is just starting to learn I wouldn't suggest "Royal Flush" as their first tournament entry. :-)

At the end of the day, I agree that proper analysis should be incorporated into playtesting. I think it is a major flaw not having it. On the other hand, TLT is not the example of mistakes I am concerned about. Pre-nerf Phantoms were, and those are something I instantly objected to, continued to rail against, and eventually started taking 86 point Phantom lists to tournaments in order to hammer my point home until changes were made. Finally, I am not anywhere near the most vocal person involved, so there is that as well. :-)

Edited by KineticOperator

A reasonably large (but not crippling) variance between ships in this game is not only completely unimportant to many of the consumers of this game, it is fantastically less varied than many flagship minis games of times past. Veteran minis war gamers find this game to be exceptionally well balanced, and the oddities are nothing compared to the outliers of many other popular games.

That's not to say there's not imbalance, or that it doesn't matter, it's just not required for a good game, or even a good competitive game. Excellent balance would be nice, but it's not really making the game a more fun, or more interesting game than the one we have now. I mean, there's a whole class of players who like to figure out uncommon, funky builds.

I see your points Juggler, but I disagree that this is a case of confirmation bias. You generally look at things in a straight line, jousting formula, but the game is NOT a straight line, dice rolling joust (at least not entirely).

I'm not using high level jousting math or averages though, I'm running the exact probabilities of what actually happened in that game. The numbers simply are what they are.

Matt took TLs, because he self bumped his Y-Wing in order to keep Jax where he was and deny the more likely 2 bank by Jax. ....

TLs were the correct choice, because they could be left for a later turn when stacked actions would be necessary to penetrate Fel's Stealth device.

Or he could have taken the TLs on Jax and used them in a later round anyway. In effect he decided to shift from going after Jax to going after Fel. In any event I ran the numbers with the actions as taken in game as as baseline as well.

In the real game, a munitions failure made it very unlikely that Jax would go down. That is another point lost in the averages.

I'm not calculating the averages, I'm calculating the exact damage PDFs. The chance of Jax dying through the end of the second combat round (round 3) if one of the Y's has munitions failure was not "very unlikely", it was 62.9%. And the chance of Jax getting away like he did that round was only 16%. And that was WITH the munitions failure, which in and of itself only has a ~2.5% chance of occurring from Turr's attack.

Analyzing the game as the actions played out, where he only focused on one Y-wing that had a shot on Jax, then the chances of killing Jax through round 2 drop slightly to:

  • with munitions failure: 62.9% (3.42 average damage, capping damage at 4)
  • without munitions failure: 71.8% (3.57 average damage, capping damage at 4)

Likewise, the chance of Jax having 2+ hull remaining (as he did in the actual match) after the 2nd round (if Matt target locks Fel) is:

  • with munitions failure: 16.0%
  • without munitions failure: 11.2%

I am not intending to deny the math, though I do contend that saying dice favored me "heavily" in the first two turns when in reality there was only a single point of damage swing between expected incoming damage and actual incoming damage is misleading.

By the end of round 3 (2nd round of actual firing) there should have been 2 more damage on Jax, who should have been dead -- and this is with "neutral" luck. If it had equivalently gone the other way, then Jax would have been dead and it would have been 4 shielded Y-wings vs 2 Interceptors. The numbers are what they are.

The only debate is whether or not the start of round 4 would have been meaningfully different if Jax was dead and all 4 Y-wings still had some shields left, vs what actually happened in the game. I personally think that's a significant difference.

Saying that "if the munitions failure had not happened" is also misleading, because lopsided vulnerability to outliers/crits is one of the greatest disadvantages TLT lists share.

It's extremely meaningful, because I did not run the exact numbers on how lucky you got in your attack. If we include the probability that you get a munitions failure on your attack then luck favors you even more than the numbers above indicate.

A fair analysis of any matchup is to re-run the dice rolls 2 ways: once with "neutral" dice, and again with the luck exactly opposite of what happened in the game. Comparing these two with the actual game tape gives the best unbiased comparison of how the match would play out if repeated: luck neutral, luck pro one player, and luck pro the other player. Alternativly, you can play the match 12 times and see what happens. But extrapolating results from one match without accounting for skewed luck is going to lead to wrong conclusions.

Mind you, 3 Aces may still have the edge vs TLT, but you can't make that conclusion from that one match.

At the end of the day, I agree that proper analysis should be incorporated into playtesting. I think it is a major flaw not having it. On the other hand, TLT is not the example of mistakes I am concerned about. Pre-nerf Phantoms were, and those are something I instantly objected to, continued to rail against, and eventually started taking 86 point Phantom lists to tournaments in order to hammer my point home until changes were made. Finally, I am not anywhere near the most vocal person involved, so there is that as well. :-)

Playtesting should not be unduly affected by how vocal someone is, but again this speaks more to FFG's playtesting weakness.

Thanks for this post, KO! Very good!

I think you missed my point. Stating that "if the game had played out exactly the same way over those 3 rounds but the dice had been different" is not really valid, because if the dice had been different the decisions made would also have changed. It is especially difficult to say this because those lists that are most capable of adapting "on the fly" to dice results are also those that are least efficient in terms of pure jousting.

At least, now they are. Pre-nerf Phantoms were both extremely efficient AND the most adaptable ships in the game.

I understand that you ran the numbers for this particular game, but your margin of error increases dramatically with every round, because decisions predicated on game-state are being made as you go along. The most accurate round is round 1, where the damage variance was very low. The real result was one of the two most likely game states (2 or 3 damage). Round 2 with a 1-hull Jax sees me putting all 3 ships at point blank on the wounded Y-Wing, he is unlikely to have survived at all regardless of Munitions Failure. I also had the option of simply declining the engagement, turning to put Jax out of range of his 3 lead ships entirely. What you are calculating is not whether or not Jax survives for 2 or 3 rounds, you are calculating the odds assuming both Matt and I had ignored the evolving game state, made exactly the same decisions, but had done so despite different results. Those are not safe assumptions.

Putting it in simpler terms by way of analogy, there is a big difference between calculating whether a ship takes (hit, hit, crit) in that order, or calculating the odds that he takes 2 hits and a crit in any order. What you are doing is functionally similar to the former, and it's accuracy for analyzing an unfolding game state decreases with every decision made.

Basically, you can either look at all the dice through a game and draw some generalized conclusions about dice skew, or look at dice round by round to see if early outliers were critical. If you do the latter, you need to keep in mind that your confidence decreases dramatically with each round because decisions are made based on earlier results. If you do the former, you have to keep in mind that early outlier results are far more influential than later ones.

I know you disagree with me (and Matt) about the TL decision, but I felt at the time and still do feel that it was the right call. TLs on a ship you are very likely to kill are wasted actions, because they go away. Focus on ships that are likely to get "Jax'd" are also wasted actions. When you are unsure about what targets you will have and whether or not your foucs will get "Jax'd" then grabbing TLs on a ship you know you will need action stacks against eventually is the right play, IMO. More to the point, if those ships had TLs on Jax, Fel would have barrel rolled right into their face and put an additional 4 dice into the lead ship rather than take the 3v2 long range poke (that dealt 0 damage). The greatest advantage the Interceptors have is encapsulated right here. Matt had to make decisions based on worst case scenarios, I had the luxury of complete intelligence when making my decisions and could make the optimal one.

Edited by KineticOperator

4x Ys with ICT, BTL-A4, and R4 agromechs do really dirty things to 4 TLT

I understand -- in practical terms this means that you are able to better leverage your early (mild) luck into the next round. If Jax has 2 hull left, then move A is optimal, and backup move B is not needed. Move B is normally inferior to move A, otherwise move B becomes the default move in all cases.

If you are forced into move B because you had average or sub-par luck, that your position in the next round suffers. This is exactly the kind of snowball effect that I was talking about. If you take the extra damage then you are worse off both position AND dice value.

I think you are trying to argue it both ways and say that 1 hull Jax + move B would be an even better position than 2 Hull Jax + move A. But that doesn't make any sense to me, because then that means move B is universally better, which if true means you should have just done that move to begin with anyway. I'm probably missing something, but in this case, it looks like putting the ships all in range of the wounded Y would have involved turning Fel in closer -- Turr was already there, and I'm not sure Jax could get there because he was blocked. I think Fel would have been left at R1 and R2 in-arc of the other two Y's as a result.

I think we both saying the same thing, that early changes in the luck drastically affects how the game plays out. I think we also both agree that if a player gets bad luck, it also negatively affects his viable tactical options as well, which won't directly show up in a numerical analysis of the match as played.

Kinetic, do you have advice for flying against/building against TLT stresshogs?

mxlm-

Lol, yes. Don't get stressed. :-)

Jokes aside, I made a big mistake in my game and forgot about R3A2. Mostly, just anticipate taking an early stress from them, and don't let yourself be in a position where you are either double stressed or forced into a bad play to clear stress. The reason I put Turr in front is because he is good at flying without necessarily using PtL, but then I forgot and did it anyway.

A Stress-HOG on the other hand is an absolutely lethal ship that has to be avoided at range 1-2 at nearly all costs, and focused down at the earliest opportunity even if it isn't otherwise optimum. Get past him, control him if you have elements for that, but otherwise just put him at the very top of your target priority.

Juggler -

I agree we are mostly saying the same thing, but I was a bit unclear about why Interceptors are particularly good in this environment. Interceptors can take one of their own that is wounded by early outliers and "hide" it, using it as bait to force low-odds shots or a switch to a different target. They can also pounce on a ship that has been wounded by early outliers and finish it off quickly. Y-Wings and TLTs are not really able to control the game-state in this way. They have very limited options in terms of adapting through maneuver, and the nature of TLT damage makes it far more difficult to achieve a beneficial outlier result in the first place.

If it were TLTs vs. TLTs, then you would be much more likely get the "snowball" you are talking about because there is much less that can be done to alter the raw numbers.

Edited by KineticOperator

I agree we are mostly saying the same thing, but I was a bit unclear about why Interceptors are particularly good in this environment. Interceptors can take one of their own that is wounded by early outliers and "hide" it, using it as bait to force low-odds shots or a switch to a different target. They can also pounce on a ship that has been wounded by early outliers and finish it off quickly. Y-Wings and TLTs are not really able to control the game-state in this way. They have very limited options in terms of adapting through maneuver, and the nature of TLT damage makes it far more difficult to achieve a beneficial outlier result in the first place.

If it were TLTs vs. TLTs, then you would be much more likely get the "snowball" you are talking about because there is much less that can be done to alter the raw numbers.

Forcing a ship to bug out of the fight and not inflict damage on you helps alter the raw numbers though! :)

I would add, also, that in the KO's TC Aces match of 3 Interceptor Aces vs TLTs, he had extraordinarily good luck

MJ, I'm sure you are proud of your past victories (as you should be!), so consider that other people might be proud of theirs as well before feeling like you need to tell them that they should have lost.

But this thread has gotten off track enough, so I'll stop here...

KO, thanks for the thoughtful suggestions for taking down TLTs. You gave a great starting point for a lot of players. I am playing around with phantoms as a TLT counter.

Echo with SJ, rec spec, LW, ACD

Sigma with IA, Stygium

OGP with Palpatine

Echo avoids tons of damage and dances around at range 3, the Sigma rushes into range 1 to block, and the emperor runs away. I took this list to a 15 person tournament yesterday, and won*. Unfortunately, I didn't get to test it against a TLT spam list.

*Edit: I figured I would add this in myself, because on two occasions, I rolled 4 natural evades with Echo.

Edited by Criwi Romed