[Joe Willie]

I am looking for ways to Rank Players. I currently run two tournments. One monthly and one Quarterly. I currently use just a basic ranking of 1 point for each place. If there are 40 people in the tournament the winner gets 40 pts. I have been playing around with some other methods that increase the possible pt totals based on the number of plays therefore accounting for the difficultly of larger tourneys. Also this helps if you miss a low entry tourney but won a large entry one.

I would be interested in any methods that others have tried.

[dammitjim]

We've been ranking based on # of opponents beat:

# of opponents beaten/# of opponents faced * 200 = a score where 100 means you've defeated 1/2 of the opponents you've faced.

We also have a quarterly tourney where the starting chip counts are based on your performance in the previous tourneys. It ends up emulating a final table with big and small stacks. We've been expreimenting with an escalating players beaten=chips for the tourney system.

Basicly say for the first 10 players you bet you get 5k each for the tourney. Then players 11-15 get you 7.5k each. 16-20 15k each, etc. It's not exactly like that, but you get the idea.

[omar]

I have a home game with 7-10 players. We are all putting $3/tourney into a TofC tourney at the end of the quarter (I got the idea fromthe CT Thursday tourneys). TofC points are awarded relative your finsih in the regular tourneys, 1 point for finishing last, 2 for next to last, etc. However, finishing in the money also gets bonus points, for example if we have 8 players and pay out 2 spots, then 3rd place gets 6 points, second pplace gets 7 plus 1 bonus point an first gets 8 plus 2 bonus points.

I am not sure if this is relative your question or not, but just thought I would throw it out there.

[Dimes]

When you're 40 players, I think the simple points structure

40-39-38-37-36-35-34-33-32-31-30 etc.

is too flat when you get over a handful of players. The winner in this example only gets slightly more (2.6%) than number two.

Sports ratings based on giving points in a number of events in which a significant number of players participate almost always space out the points for the top finishes. The difference in points between first and second is bigger than between last and second last. Such systems are used in golf, tennis, motorsport and probably in many other sports and competitions.

Now, how do you create such a scale? There are different ways to do it. The simplest way to do it is to add some bonus points for finishing in the best positions. If you are 40 players, you could award 5 bonus points for finishing first, 3 for second and 2 for third and 1 for fourth. Then the scale would be

45-42-40-38-36-35-34-33-32-31-30 etc.

You could also award 10-6-4-2-1 bonus points to the top 5 and get

50-45-42-39-37-35-34-33-32-31-30 etc.

If you want to favour the top positions even more, you could have:

55-49-44-41-38-36-34-33-32-31-30 etc.
60-52-46-42-39-37-35-33-32-31-30 etc.

You could adjust the bonus point scale so that the players finishing in the money are eligible for bonus point.

There are some methods to create such scales in a systematic way, but I think I will leave it here for now.

TT

[SpeakEasy]

Use the formula that is used for our weekly CT tournaments, compliments of CapnLego. It awards a proportionally higher score for those that last longer in the tournament each week. Check the threads in the tournament forum.

[Joe Willie]

I looked for the CT Tourney formula but couldn't find it. Which thread is it in?


I also found this natual logarithem method has anyone used this? You can average the ranking over the tournaments played in.

(LN((# players+1)/Place finshed))

1st place finish in a 14 person tourney (LN((14+1)/1)=2.7081
2nd place finish in a 14 person tourney (LN((14+1)/2)=2.0149
3rd place finish in a 14 person tourney (LN((14+1)/3)=1.6094

1st place finish in a 40 person tourney (LN((14+1)/1)=3.7136
2nd place finish in a 40 person tourney (LN((14+1)/2)=3.0204

Does anyone know how the Toyota Rankings are calculated?

[SteveA]

Joe,

I use the same forumla as you, but I also award points if you end up in the money. For example, if 5 spots are paid out, 1st place gets an additional 5 points, 2nd 4, 3rd 3, etc..

I also do a tournament of champions where the starting stacks are proportionate to points. Check out my rules here: http://www.homepokerclub.net/clubrules.php?clubid=1

[CaptLego]

Originally Posted by Joe Willie

I looked for the CT Tourney formula but couldn't find it. Which thread is it in?

http://chiptalk.net/forum/showthread.php?t=8999
For a bit more analysis, see posts #28 and #31 in that thread.

[whitepotatoe]

If you only award points according to how many players there are you have no idea who is good and who is bad. For example, in a 40 player tourney, say you have player A and player B. Player A takes first place every third tournament, but is one of the first people to bust out 2/3 of the time. He would average about 15 points per tournament. Player B never makes the final table and is way too conservative, but he never busts out early because he will only play with the nuts. He usually busts out about in about 20th - 25th place. He averages 17 points per tournament.

Clearly player A is good and player B stinks. Why does player B have more points? Because that system does not work. Its the same reason that tournaments pay out the way they do. 1st place should get significantly more than 2nd, which should get significantly more than 3rd, etc.

If you have 40 players, you should do something like this

40th - 1
39th - 2
38th - 3
...
...
12th - 28
11th - 29
10th - 30
9th - 32
8th - 34
7th - 36
6th - 38
5th - 42
4th - 46
3rd - 53
2nd - 64
1st - 80

If you play a decent number of tounraments this will show who the best players are.

[EDIT] I just looked at what Capt has and that seems like a pretty good way to do it. Whichever way you do it, you have to weight the top places much more than the bottom ones, but it is good to give some points for participation, but not so much that a regular particapator who sucks can end up doing better than a guy who comes half the time and is awesome.

[Joe Willie]

Originally Posted by CaptLego

http://chiptalk.net/forum/showthread.php?t=8999
For a bit more analysis, see posts #28 and #31 in that thread.

Thanks for the link Capt. I agree that the the top spots (in the money) should get a few more points than the other finishers. In the long run the numbers should average themselves out so that most of these methods will show a true players ranking.

I am ponering this thought right now, if a good player only plays in the large field events ( a la Tiger Woods) his risk reward in the events he plays should level out the limited participation