Tibs kindly sent me the latest stats sheet, so here we are. I have been working for a while on a Python script that can parse the spreadsheet (in OpenDocument format, the native format for Google docs, that is also the native format for OpenOffice) and then calculate the stats. I initially focussed on the "missing stats", the ones that were difficult to do from a spreadsheet software, and then extended to almost all the stats Tibs is used to post here.

Therefore I will show the "missing" stats, in particular the expansion combos stats and the individual investigator success rates.

• Cumulated Expansion stats

There is shown the cumulated expansion usage:

Total number of games: 1057
Expansion #games #vict %games %vict %vict with this exp
Base 205 150 19.39 14.19 73.17
BGotW 84 58 7.95 5.49 69.05
CotDP 487 305 46.07 28.86 62.63
DH 631 391 59.70 36.99 61.97
KH 355 207 33.59 19.58 58.31
KiY 450 279 42.57 26.40 62.00

• Expansion Combination stats

There is the expansion usage, breakdown for all possible combinations

Combo #games #vict %games %vict %vict with this combo
('',) 205 150 19.39 14.19 73.17
('BGotW',) 8 5 0.76 0.47 62.50
('CotDP',) 55 37 5.20 3.50 67.27
('DH',) 152 93 14.38 8.80 61.18
('KH',) 48 27 4.54 2.55 56.25
('KiY',) 34 25 3.22 2.37 73.53
('BGotW', 'CotDP') 3 2 0.28 0.19 66.67
('BGotW', 'DH') 3 3 0.28 0.28 100.00
('BGotW', 'KH') 1 1 0.09 0.09 100.00
('BGotW', 'KiY') 1 1 0.09 0.09 100.00
('CotDP', 'DH') 72 53 6.81 5.01 73.61
('CotDP', 'KH') 13 11 1.23 1.04 84.62
('CotDP', 'KiY') 14 7 1.32 0.66 50.00
('DH', 'KH') 33 19 3.12 1.80 57.58
('DH', 'KiY') 51 31 4.82 2.93 60.78
('KH', 'KiY') 24 19 2.27 1.80 79.17
('BGotW', 'CotDP', 'DH') 2 0 0.19 0.00 0.00
('BGotW', 'CotDP', 'KH') 3 1 0.28 0.09 33.33
('BGotW', 'CotDP', 'KiY') 2 2 0.19 0.19 100.00
('BGotW', 'DH', 'KH') 0 0 0.00 0.00
('BGotW', 'DH', 'KiY') 1 1 0.09 0.09 100.00
('BGotW', 'KH', 'KiY') 0 0 0.00 0.00
('CotDP', 'DH', 'KH') 8 6 0.76 0.57 75.00
('CotDP', 'DH', 'KiY') 63 43 5.96 4.07 68.25
('CotDP', 'KH', 'KiY') 15 8 1.42 0.76 53.33
('DH', 'KH', 'KiY') 9 7 0.85 0.66 77.78
('BGotW', 'CotDP', 'DH', 'KH') 1 0 0.09 0.00 0.00
('BGotW', 'CotDP', 'DH', 'KiY') 36 27 3.41 2.55 75.00
('BGotW', 'CotDP', 'KH', 'KiY') 0 0 0.00 0.00
('BGotW', 'DH', 'KH', 'KiY') 0 0 0.00 0.00
('CotDP', 'DH', 'KH', 'KiY') 177 93 16.75 8.80 52.54
('BGotW', 'CotDP', 'DH', 'KH', 'KiY') 23 15 2.18 1.42 65.22

I'll post comments on these stats later.

Just looking at significant stats for combos (where there are at least 20 games played with that combo):

('',) 205 150 19.39 14.19 73.17

('BGotW',) 8 5 0.76 0.47 62.50
('CotDP',) 55 37 5.20 3.50 67.27
('DH',) 152 93 14.38 8.80 61.18
('KH',) 48 27 4.54 2.55 56.25
('KiY',) 34 25 3.22 2.37 73.53

('CotDP', 'DH') 72 53 6.81 5.01 73.61

('DH', 'KH') 33 19 3.12 1.80 57.58
('DH', 'KiY') 51 31 4.82 2.93 60.78
('KH', 'KiY') 24 19 2.27 1.80 79.17

('CotDP', 'DH', 'KiY') 63 43 5.96 4.07 68.25

('BGotW', 'CotDP', 'DH', 'KiY') 36 27 3.41 2.55 75.00

('CotDP', 'DH', 'KH', 'KiY') 177 93 16.75 8.80 52.54
('BGotW', 'CotDP', 'DH', 'KH', 'KiY') 23 15 2.18 1.42 65.22

Unsurprisingly, adding an expansion to the base game increases the difficulty (in particular DH and KH), except KiY.

But there are surprises: ('CotDP', 'DH') is more successful than DH, meaning that CotDP makes DH easier than DH alone! This is confirmed by ('CotDP', 'DH', 'KiY') as well. Maybe only experienced players play these multiple expansions? On the other hand ('CotDP', 'DH', 'KH', 'KiY') is the hardest combo! Strange, isn't it ? Of course, we need to be careful when the number of games is low, because the statistical variations will be higher.

In term of number of games, it is natural to see combo frequency follow the order of expansion publishing date. But there are some peculiarities: there are few "out-of-order of publication" games, such as DH + KiY or KH + KiY. There are more CotDP+DH, CotDP+DH+KiY, etc which means people don't bother separating their extensions after purchase. For instance, there are more ('CotDP', 'DH', 'KH', 'KiY') than any other combo, even DH alone! It means that people "accumulate" their expansions and they do not remove them afterwards.

• Investigator popularity and individual success

This table shows the % of games and % of victories for each investigator:

Name                 #games     #vict      %games     %vict      %vict with this inv"Ashcan" Pete          118         88      11.16       8.33      74.58Amanda Sharpe           81         61       7.66       5.77      75.31Bob Jenkins             96         69       9.08       6.53      71.88Carolyn Fern           118         85      11.16       8.04      72.03Charlie Kane            48         24       4.54       2.27      50.00Daisy Walker            79         48       7.47       4.54      60.76Darrell Simmons        137        105      12.96       9.93      76.64Dexter Drake           128         83      12.11       7.85      64.84Diana Stanley          104         69       9.84       6.53      66.35Gloria Goldberg         92         61       8.70       5.77      66.30Harvey Walters         103         67       9.74       6.34      65.05Jacqueline Fine        104         80       9.84       7.57      76.92Jenny Barnes           109         72      10.31       6.81      66.06Jim Culver              70         40       6.62       3.78      57.14Joe Diamond            129         93      12.20       8.80      72.09Kate Winthrop          113         81      10.69       7.66      71.68Leo Anderson            99         67       9.37       6.34      67.68Lily Chen               84         55       7.95       5.20      65.48Lola Hayes              56         39       5.30       3.69      69.64Luke Robinson           58         38       5.49       3.60      65.52Mandy Thompson         202        151      19.11      14.29      74.75Marie Lambeau           74         48       7.00       4.54      64.86Mark Harrigan           88         62       8.33       5.87      70.45Michael McGlen         116         75      10.97       7.10      64.66Monterey Jack          106         71      10.03       6.72      66.98Rex Murphy              58         34       5.49       3.22      58.62Rita Young              86         60       8.14       5.68      69.77Sister Mary             85         43       8.04       4.07      50.59Tony Morgan             57         37       5.39       3.50      64.91Vincent Lee             82         42       7.76       3.97      51.22Wendy Adams             65         47       6.15       4.45      72.31Wilson Richards         97         67       9.18       6.34      69.07

I let you draw your own conclusions and comment them.

If some people have ideas of other stats, I'll be pleased to add them to my script (as long as they are not overwhelmingly too complex).

zeb said:

Am I reading it right, 36 games with said combo, 27 wins? Hehe, I think I contributed quite a bit then, as my last 24 games were using that combo and only 2 losses.

zeb said:

Haha, KiY makes base game easier, lol.

zeb said:

I can only speak for myself, but I'm way too lazy to start separating expansions and then mixing them in again, making sure they are shuffled properly . Seems kinda redundant.

The three worse investigators in the game are Vincent, Sister and Charlie Kane (as popularly played). I've always considered them a drag on a team. Time to reimplement my house rules to make them more playable.

Vincent can heal anyone in town if he's in town too. Sister doesn't lose her blessing and Charlie (well I won't mention what I do with Charlie).

Daisy Walker at 60% probably means most people haven't discovered her spell yet.

I'm kind of surprise to see Amanda and Jacqueline so high in helping their teams. Maybe not Jacqueline. She really saved us a lot of Mythos trouble in our last game.

To Zeb: The stat I always extrapolate is the chance of winning by seals/close vs. battle for each AO.

mageith said:

I think people who draw Mary just "give up" and play her poorly, because they hate her. For me, Mary's always good .

Also, WTF is up with Amanda Sharpe and her 75+% ? No effing way! Or maybe I just play her poorly 'cos I hate her .

mageith said:

I think that Tibs already provides the chances of winning during Final Combat for each AO. But I can work out the other ways to win.

zeb said:

I use that stat to extrapolate my figures. What I am interested in the chances of winning by seals/closing when entering into a game with a particular GOO. If you look up higher in the thread you'll see my extrapolation. My extrapolation is not exact: It assumes every game is either a seal/close victory, a final battle victory or a final battle loss. Other wins and losses would be interesting.

I've included chances to win by Final Combat, and chances to win without the AO awakening (The entry on my last report is Most Restless ). This is essentially a combination of chances to seal and chances to close.

I didn't bother to differentiate between winning by closing vs. winning by sealing, because their relative chances are probably not based on the AO so much as the number of players and overall luck of the game. Only Hastur, Yibb, and Atlach should yield a significantly different ratio, but it might be worth investigating for anyone who cares.

Would it be possible to add a field for Scenarios? Especially with a 'Custom' entry, that would be really helpful.

• Stats report for 02/17/09:
_____________________

• Introduction

There have been 1119 reported sessions:

65.2% were reported as victories:

356 Victories by seals
238 Victories by Final Combat
135 Victories by closing all gates
1 Victory by "For the Greater Good" mission

...Leaving 34.8% as Defeats:

361 Defeats by Final Combat
12 Forfeits
7 Pre-Final Combat Shudde M'ell Earthquakes
6 Defeats by King in Yellow's Act III
3 Successful completions of "Joining the Winning Team"

_____________________


• Ancient Ones

• Most popular
Here are the Ancient Ones, in order from most to least commonly played, followed by the number of games:

1. Nyarlathotep 116
2. Yig 105
3. Hastur 94
4. Cthulhu 90
5. + Shub-Niggurath 88
6. - Azathoth 82
7. ++ Yog-Sothoth 72
8. - Ithaqua 70
9. - Shudde M'ell 67
10. + Abhoth 62
11. - Tsathoggua 62
12. Glaaki 55
13. Y'Golonac 49
14. Atlach-Nacha 42
15. Yibb-Tstll 34
16. Eihort 31

• Most dangerous
Here are the Ancient Ones, from highest to lowest investigator defeat percentage:

1. Atlach-Nacha 59.5%
2. Y'Golonac 53.1%
3. Glaaki 50.9%
4. Abhoth 43.5%
5. + Shudde M'ell 40.3%
6. - Cthulhu 40.0%
7. ++ Tsathoggua 38.7%
8. - Yibb-Tstll 38.2%
9. - Shub-Niggurath 37.5%
10. + Azathoth 32.9%
11. + Eihort 32.3%
12. -- Yog-Sothoth 30.6%
13. Hastur 28.7%
14. Ithaqua 25.7%
15. Yig 21.9%
16. Nyarlathotep 19.8%

• Most restless
Here are the Ancient Ones, from highest to lowest percent of games where it has awoken (this is essentially the difficulty to close or seal):

1. Yig 80.0%
2. Glaaki 69.1%
3. + Atlach-Nacha 69.0%
4. + Y'Golonac 65.3%
5. -- Abhoth 64.5%
6. Shudde M'ell 50.7%
7. + Hastur 58.5%
8. + Shub-Niggurath 55.7%
9. -- Nyarlathotep 55.2%
10. + Ithaqua 50.0%
11. - Yibb-Tstll 47.1%
12. Cthulhu 43.3%
13. + Eihort 41.9%
14. - Yog-Sothoth 37.5%
15. Tsathoggua 33.9%
16. Azathoth 28.0%

*This includes the investigators' losses by Shudde M'ell wrecking the town.

• Hardest to fight
Here are the AOs, from highest to lowest investigator defeat percentage during Final Combat:

1. Azathoth 100.0%
2. Tsathoggua 95.2%
3. Atlach-Nacha 86.2%
4. Cthulhu 84.6%
5. Yog-Sothoth 81.5%
6. Y'Golonac 81.3%
7. Yibb-Tstll 81.3%
8. Eihort 76.9%
9. Glaaki 68.4%
10. Shub-Niggurath 67.3%
11. Abhoth 62.5%
12. Shudde M'ell 58.8%
13. Ithaqua 51.4%
14. Hastur 43.6%
15. Nyarlathotep 32.8%
16. Yig 26.2%

_____________________

• Heralds

• Most popular
In games that used a Herald, the percent of those games that each Herald was used:

1. King in Yellow 31.1%
2. Tulzscha 17.2%
3. + Ghroth 16.1%
4. - Dunwich Horror 15.5%
5. Dark Pharaoh 13.5%
6. Black Goat 6.6%

59.5% of games did not use a herald.

• Herald effectiveness
The percent of investigator-lost games, by herald:

1. + Ghroth 45.2%
2. - Dunwich Horror 44.3%
3. Tulzscha 38.5%
4. King in Yellow 34.8%
5. No herald 32.7%
6. Dark Pharaoh 31.1%
7. Black Goat 30.0%

• Unattempted combos
These are combinations of Ancient Ones and Heralds that have not yet been attempted:

• Atlach-Nacha / Black Goat
• Yibb-Tstll / Black Goat

Only one has been attempted since the last report. Get to it, investigators!!!


• Undefeated combos
These are combinations of Ancient Ones and Heralds that have been attempted at least once, but have not been won by the investigators (combos still undefeated from the last report are in bold):

• Abhoth / Black Goat
• Y'Golonac / Black Goat
• Yog-Sothoth / Black Goat
• Y'Golonac / Dark Pharaoh
• Atlach-Nacha / Dunwich Horror
• Y'Golonac / Dunwich Horror
• Glaaki / Tulzscha

Congrats to Acebob for toppling Yog-Sothoth / Ghroth and to Bel-Shamharoth for defeating both Eihort / Dark Pharaoh AND Shub-Niggurath / Ghroth !


_____________________

• Guardians

• Most popular
In games that used a Guardian, the percent of those games that each Guardian was used:

1. Nodens 40.2%
2. Hypnos 34.2%
3. Bast 25.6%

71.8% of games did not use a guardian.

• Guardian effectiveness
The percent of games investigators won, by Guardian:

1. No guardian 67.6%
2. Nodens 59.8%
3. Bast 59.3%
4. Hypnos 58.3%

Oh cool. The heralds have all come very close to offering equal protection.

_____________________

• Investigator Popularity
Investigators sorted from most games used to least:

1. Mandy Thompson 215
2. Darrell Simmons 151
3. Joe Diamond 139
4. Dexter Drake 135
5. Carolyn Fern 131
6. "Ashcan" Pete 131
7. ++ Jenny Barnes 127
8. - Michael McGlen 126
9. - Kate Winthrop 117
10. +++ Harvey Walters 113
11. - Monterey Jack 110
12. - Diana Stanley 110
13. +++ Leo Anderson 109
14. -- Jacqueline Fine 107
15. - Bob Jenkins 102
16. - Wilson Richards 101
17. Gloria Goldberg 99
18. Mark Harrigan 95
19. Sister Mary 93
20. +++ Amanda Sharpe 93
21. - Rita Young 91
22. - Vincent Lee 89
23. - Lily Chen 88
24. Daisy Walker 84
25. Marie Lambeau 81
26. Jim Culver 76
27. Wendy Adams 69
28. Rex Murphy 64
29. + Luke Robinson 62
30. - Lola Hayes 61
31. Tony Morgan 58
32. Charlie Kane 49

• Team numbers
How many games involve how many investigators:

1 26
2 180
3 229
4 523
5 84
6 56
7 13
8 7
9 1


• Team size vs. Chance to win
Percent of games won by certain sized investigator teams:

1 46.2%
2 58.9%
3 61.1%
4 70.4%
5 69.0%
6 62.5%
7 46.2%
8 42.9%
9 not enough data


• Team size vs. % won by closing all gates:

1 7.7%
2 4.4%
3 9.2%
4 14.1%
5 21.4%
6 16.1%
7 15.4%
8 not enough data
9 not enough data


• Team size vs. % won by sealing gates:

1 3.8%
2 25.6%
3 26.2%
4 37.7%
5 29.8%
6 41.1%
7 15.4%
8 not enough data
9 not enough data

• Team size vs. % won by Final Combat
If the Ancient One awoke, the percent of games where teams of a certain size were victorious in Final Combat.

1 40.9%
2 43.0%
3 42.8%
4 40.4%
5 36.6%
6 13.0%
7 22.2%
8 20.0%
9 N/A


• Investigator selection method:

Random 71.8%
Before AO 22.2%
After AO 5.9%

_____________________

• Expansions used
The percent of games that used any given expansion:

DH 59.5%
CotDP 44.7%
KiY 42.5%
KH 32.9%
none 20.4%
BGotW 8.4%


• Special events
The number of games in which a given special event occured

Dual-Color OW battle 32
"Old Debt" paid to the Dark Man at Wizard's Hill 3
Coded Messages activated 7
Darke's Blessing activated 4
Rare Book Collection activated 1
Velma's Gratitude activated 4
DH awoke 137 (20.6% of all games using the Dunwich board)
DH awoke w/o DH herald 101 16.9%
DH awoke with DH herald 36 51.4%
% increase from using he DH herald 203.5%
DH was defeated 34 (24.8% of games in which it awoke)
Terror Level reached 10 75
At least one rift opened 19
"Switched bodies" at the Science Building 2
Devoured by 0 San/Sta simultaneously 6
A gate burst removed a seal 28


• Special acquisitions
The number of games in which a particular title was gained

Anointed 4
Barred from a Neighborhood 16
Beloved of Bast 6
Captain of the White Ship 27
Changed 17
Deputy of Arkham 212
Harried 5
Local Guide 6
One of the Thousand Cult Membership 22
Private Investigator 2
Psychic 9
Rail Pass 12
Sheldon Gang Membership 18
Silver Twilight Lodge Membership 173
Tainted 6
Visions 15
Wanted 2


• Ways the Ancient One awoke
Number of reported games where the AO awoke a given way:

The doom track filled 78
Too many gates were open 30
Too many monsters (after TL10) 11
Ran out of monster tokens 1
Ran out of Servants of Glaaki 1
Call Ancient One spell 2

_____________________

• Top contributors
These are the individuals who contributed the largest number of games:

1. Tibs 93 8.3%
2. mageith 87 7.8%
3. Dam 72 6.4%
4. jhaelen 44 3.9%
5. flamethrower49 36 3.2%
6. Victimizer 33 2.9%
7. Cariosus 29 2.6%
8. Morgaln 24 2.1%
9. Vemb 23 2.1%
10. dj 22 2.0%

I think flamethrower49 took everyone's recommendations about expansions seriously. All of his submissions have been with the Dunwich Horror and King in Yellow expansions. He's also climbing the ladder like a maniac.

I love these.

I only wish I had time to do the league and attempt the unbeaten combinations!

Is there a good way to get comprehensive rumour statistics? Like which rumours contribute to the success or failure of the investigators? That sounds interesting.

awp832 said:

Last time I updated the forum, I entered checkboxes for passed or failed rumors. The best I can do for now is wait for a significant number of entries, and then reveal how often each rumor is passed or failed. It would be really difficult to correlate how these affected the entire game's outcome.

• Stats report for 03/02/09:

(Deleted: obsolete)

Kudos again, Tibs. Very interesting stuff. A suggestion though, flamethrower49 and myself agree that we'd be very interested to see a new column of % games won-lost with certain Rumors passed/failed. Kind of a 'deadliest rumors' kind of thing. Also, #4 baby! yeah! Dam, we're gunning for you!

1193 games. 32.9% used Kingsport. Only 19 games had any rifts open. That's less than 5%. I had 4 of them and Tibs had more than me.

Interesting.

It's interesting to me that nearly 1/3rd of games use Kingsport and that the rifts are so fearsome that we keep them very well undercontrol.

Almost as many have been Changed (17) as have seen a rift in action.

Tibs Thanks for the reports. I look forward to them!

awp832 said:

Hehe, as I've said before, my last game was in mid-December. Also, doesn't look like I'll be going from 72 to 84 (7th rotation of AH + DH GOOs) anytime soon, Age of Conan should be in stores this month, looking to get Lord of the Rings TBG as well, maybe May will bring a change with Innsmouth, although the Dungeon expansion for Talisman is also scheduled for May, so who knows. Most likely it'll be summer when I next crack open my AH .

Well guys, I have responses for you--

flamethrower49 (and friend):

You can see my stats submission page includes rumors. I wanted to wait for few stats reports before reporting on the rumors, so that I can gather up a bit of info.

Mageith:

I put the exact number of instances of open rifts and other events on the list instead of % of games for a reason. Those are the number of reported instances of opening rifts since I added that option to the stats page; it's not meant to be turned into a percentage. Every one of the submissions I've personally made was after I'd gotten Kingsport, but the "rift opened" check box was added fairly recent.

• Stats report for 03/16/2009:

(Deleted: obsolete)

Sweet. I didn't realize that Yog-Sototh/Black Goat was on the list. Good for us!

I think it's interesting that the still-undeafeated combos seem to be people looking at the Kingsport Ancient ones and the Black Goat and saying, "Nuh-uh." I realize that most of us play random, as well, but I think I'd be tempted to try it, if I owned Kingsport.

flamethrower49 said:

Kingsport is currently being translated into German. I really hope it gets published soon - I want to try all of those new combos!

Tibs wrote: "• Stats report for 03/16/2009:

• Special events
At least one rift opened 28 (35.4% of all games using the Kingsport board)"

Thanks again for the report.

All games? or is this the percent of games since you started recording it?

mageith said:

Since recording that feature, so it should be pretty accurate. I implemented that percentage just for you