Thoughts on Group Knockout/Strikes format w/ 3 strikes per group?


#82

Good to know. Thanks josh.


#83

Oh. Well I guess I owe Adam and email about that progressive strike tournament I just submitted.


#84

OK, so I think it’s awesome people are starting to toy with this format and am excited there are some upcoming events.

I was interested to get some idea of how long these things would run, so I just wrote a program to do it. I’ll even try to upload it so you can run it on a windows command prompt.

I called it “FairStrikeSim” as much as I also like Keefer Strikes @haugstrup. :wink:

Here are the results for some of the more interesting combinations of players and strikes (10,000 simulations each)

players strikes average lasted players strikes average lasted
8 5 6.6848 8 6 7.8153
9 5 6.9656 9 6 8.1379
10 5 7.0884 10 6 8.2603
11 5 7.1888 11 6 8.3849
12 5 7.2384 12 6 8.4413
13 5 7.486 13 6 8.6784
14 5 7.5577 14 6 8.731
15 5 7.5965 15 6 8.7948
16 5 7.6032 16 6 8.8291
20 5 7.8798 20 6 9.1227
24 5 8.2071 24 6 9.4062
28 5 8.3265 28 6 9.5178
32 5 8.4154 32 6 9.6531
40 5 8.6648 40 6 10.0469
50 5 9.0913 50 6 10.3138
64 5 9.2425 64 6 10.4742
75 5 9.3307 75 6 10.7284
100 5 9.7742 100 6 11.1504

Wow, that pasted from excel amazingly well. Thanks, discourse!

OK, not surprisingly an .exe is not allowed, so get it here if you like: https://drive.google.com/open?id=1qqpEtnGnRb8aQrncihhwnc8dJUAHdRB8

Just tell it the number of players, number of strikes, and number of simulations to run. 10,000 seems like a pretty reasonable number to use.

I’ll put up the source when I clean it up a bit. As stated it’s natively a win64 console app, but there’s probably <10 lines to change to get it to run under gcc or something. It would also be pretty trivial to change it to some other scheme like @BonusLord’s endurance thingy.

There’s probably some mathy pattern here, but I’ll let others better qualified to figure that out.

Also, I did repeated runs on a couple of these, the data seems extremely consistent through 2 decimal places.

Additional bonus edit: Sample output from a run of 100 players/6 strkes:

**** PLAYERS PER ROUND, SUMMARY AFTER RUN 10000 ****

round 1: 100.000000
round 2: 100.000000
round 3: 100.000000
round 4: 98.433403
round 5: 85.946602
round 6: 62.916302
round 7: 39.154999
round 8: 21.553400
round 9: 11.138100
round 10: 5.203600
round 11: 2.541200
round 12: 0.332000
round 13: 0.002800
round 14: 0.000000

**** ROUNDS MATCH LASTED SUMMARY AFTER RUN 10000 ****

round 1: 0
round 2: 0
round 3: 0
round 4: 0
round 5: 0
round 6: 0
round 7: 0
round 8: 0
round 9: 0
round 10: 196
round 11: 8147
round 12: 1643
round 13: 14

**** AVERAGE ROUNDS LASTED, SUMMARY AFTER RUN 10000 ****

average match lasted 11.147500


#85

I played in my first 0112 knockout on Tuesday. 16 players and it went 8 rounds, which timed out to about 4 hours with a couple long playing moderns in the mix. I thought it felt pretty fair.

MP link for those interested - https://matchplay.events/live/xr35w


#86

How many strikes?


#87

Right, the most important variable that I forgot to mention. 6 strikes


#88

Anyone want to guess how long a 24 strike progressive knockout will take? That’s what’s going down at Trent’s barn on Saturday. Guessing there may be 50-60 players on a mix of classics and moderns. Starts at 10am and could end up being a very long haul.

Follow along here if you’re curious. https://matchplay.events/live/3r1d1

I think there will also be some streaming, but I don’t have those details.


#89

Arcade super awesome plans to stream it. So look them up on twitch. Not sure the timing of it though.


#90

How do you not go 27 outs and make it the World Series of Pinball format?

That’s like running a marathon and stopping at mile marker 25 :slight_smile:


#91

I mean, it could be 24 strikes for the home team if they’re winning in the bottom of the 9th :wink:


#92

FairStrikes results:

**** PLAYERS PER ROUND, SUMMARY AFTER RUN 10000 ****

round 1: 50.000000
round 2: 50.000000
round 3: 50.000000
round 4: 50.000000
round 5: 50.000000
round 6: 50.000000
round 7: 50.000000
round 8: 50.000000
round 9: 50.000000
round 10: 50.000000
round 11: 50.000000
round 12: 50.000000
round 13: 50.000000
round 14: 49.997799
round 15: 49.978500
round 16: 49.891399
round 17: 49.630501
round 18: 49.055302
round 19: 47.934601
round 20: 46.149799
round 21: 43.339500
round 22: 39.403301
round 23: 34.213001
round 24: 28.539101
round 25: 22.702101
round 26: 17.333900
round 27: 12.652600
round 28: 8.880600
round 29: 6.009300
round 30: 3.973200
round 31: 2.505600
round 32: 1.066800
round 33: 0.195700
round 34: 0.021800
round 35: 0.001400
round 36: 0.000000

**** ROUNDS MATCH LASTED SUMMARY AFTER RUN 10000 ****

round 1: 0
round 2: 0
round 3: 0
round 4: 0
round 5: 0
round 6: 0
round 7: 0
round 8: 0
round 9: 0
round 10: 0
round 11: 0
round 12: 0
round 13: 0
round 14: 0
round 15: 0
round 16: 0
round 17: 0
round 18: 0
round 19: 0
round 20: 0
round 21: 0
round 22: 0
round 23: 0
round 24: 0
round 25: 0
round 26: 0
round 27: 0
round 28: 0
round 29: 1
round 30: 541
round 31: 4280
round 32: 4202
round 33: 867
round 34: 102
round 35: 7

**** AVERAGE ROUNDS LASTED, SUMMARY AFTER RUN 10000 ****

average match lasted 31.572701

What exactly is progressive? 0/1/2/3 0/1/2 0/1?


#93

Here’s 01 012 0123 for 50 players, 24 strikes.

Interesting to note there’s WAY more volatility with tournament length with this version of progressive.

**** PLAYERS PER ROUND, SUMMARY AFTER RUN 10000 ****

round 1: 50.000000
round 2: 50.000000
round 3: 50.000000
round 4: 50.000000
round 5: 50.000000
round 6: 50.000000
round 7: 50.000000
round 8: 50.000000
round 9: 50.000000
round 10: 50.000000
round 11: 50.000000
round 12: 49.999901
round 13: 49.954399
round 14: 48.035099
round 15: 41.346901
round 16: 33.713699
round 17: 25.916300
round 18: 18.979300
round 19: 13.074300
round 20: 9.124400
round 21: 6.171400
round 22: 4.084800
round 23: 2.553200
round 24: 1.437800
round 25: 0.819800
round 26: 0.459300
round 27: 0.252600
round 28: 0.133000
round 29: 0.073200
round 30: 0.037800
round 31: 0.018000
round 32: 0.009000
round 33: 0.003600
round 34: 0.001800
round 35: 0.001000
round 36: 0.000400

**** ROUNDS MATCH LASTED SUMMARY AFTER RUN 10000 ****

round 20: 0
round 21: 12
round 22: 697
round 23: 2565
round 24: 2647
round 25: 1783
round 26: 1033
round 27: 598
round 28: 299
round 29: 177
round 30: 99
round 31: 45
round 32: 27
round 33: 9
round 34: 4
round 35: 3
round 36: 2

**** AVERAGE ROUNDS LASTED, SUMMARY AFTER RUN 10000 ****

average match lasted 24.503201


#94

Exactly


#95

Very cool! Thanks for making this, as it really shows the equability of rounds in the 0/1/1/2 format.

I tried running the exe locally but got a number of missing DLL errors (probably my system setup, as it says it can’t find ucrtbased.dll + VCRUNTIME140D.dll).

Would you be able to run a simulation for 8 Players, 4 Strikes? We’re intending to use this as a B-Division & C-Division Final for an upcoming event as something different than the traditional PAPA-Style groups.


#96

Okay so 35-36 rounds at ~24 minutes per round is 14-15 hours. Starting at 10am that would wrap up around midnight, which is roughly what I was expecting. Thanks for the number crunching!


#97

Oh that might be because I built as debug. Let me build release and see if that helps. I’ll do the sim real quick too.

Edit for sim:

8 players, 4 fair strikes

**** PLAYERS PER ROUND, SUMMARY AFTER RUN 10000 ****

round 1: 8.000000
round 2: 8.000000
round 3: 7.498000
round 4: 5.225100
round 5: 3.126500
round 6: 1.080800
round 7: 0.015400

**** ROUNDS MATCH LASTED SUMMARY AFTER RUN 10000 ****

round 1: 0
round 2: 0
round 3: 0
round 4: 6
round 5: 4649
round 6: 5268
round 7: 77

**** AVERAGE ROUNDS LASTED, SUMMARY AFTER RUN 10000 ****

average match lasted 5.541600


#98

New links to try:

Fair Strike Sim: https://drive.google.com/open?id=12cHo_TOSrVI4Hjv7bJgJW_3bgjpy_Kkw

Progressive Strike Sim: https://drive.google.com/open?id=14bzlLHREQNebn6w1B7V1SlvKvbrW6Ac7

Feel free to try and LMK if there’s still a problem.


#99

Awesome!

I was getting DLL errors too, but this one’s working. Thanks so much for this tool!


#100

Hmmm, something’s not right. I’ve run a simulation 60 player, 7 keefer-strike tournament on Match Play that went for 16 rounds: https://matchplay.live/d3nvy

The fairstrikesim gives these results for 60 players, 7 strikes:

**** PLAYERS PER ROUND, SUMMARY AFTER RUN 1000000 ****

round 1: 60.000000
round 2: 60.000000
round 3: 60.000000
round 4: 60.000000
round 5: 57.891514
round 6: 50.024925
round 7: 37.383465
round 8: 24.413816
round 9: 14.194348
round 10: 7.641338
round 11: 3.745633
round 12: 1.406503
round 13: 0.068032
round 14: 0.000290

**** ROUNDS MATCH LASTED SUMMARY AFTER RUN 1000000 ****

round 1: 0
round 2: 0
round 3: 0
round 4: 0
round 5: 0
round 6: 0
round 7: 0
round 8: 0
round 9: 0
round 10: 0
round 11: 339903
round 12: 626083
round 13: 33869
round 14: 145


#101

Perhaps it’s a difference in how matches are seeded. I basically bunch people up in order of strikes (I guess that’s Swiss). What does matchplay do (specifically, if known)?