Tuesday, August 21, 2007

Internet Euchre Hall of Fame (Part III)

One regular at in the Advanced Lounge Five told me he thought my last post was "off" because the statistics for straight winning percentage were skewed by tournament participation in the case of some players.

Initially I thought, well, tournaments might be marginally harder to cheat at (because partners are sometimes randomly assigned and can change in some formats), so what's the big deal. A couple points do need to be considered, though:

--Tournament teams are often ranked by skill rating, so players get their easiest games first. Thus successful players will play a higher percentage of their tournament games against weaker opponents than might players who play straight ladder games.

--Tournaments can be small. If they are played at odd hours, a team might play 2-3 tournies in an hour against the same 5-7 opponents. Multiply this by a couple times a week, and the diversity of opponents might be less.

--Ladder stats for tournaments include participation in gimmick tournaments such as "no loners"; "call black only"; "reverse," etc. Generally speaking, the more variable you eliminate, the less strategy is involved.

Let's look at a view examples:

_95crownvic's ladder stats are: 6282-5081. (Winning perecentage: 55.285). I mentioned in the last post that this player has one of the highest winning percentages for any player over 1,000 games. When we factor in tournaments, however, something interesting happens.

Tourny Record (2850-1672) (Winning Percentage: 63.02)
Non-Tourny Record (3432-3409)(Winning Percentage: (50.16)

So this player's winning percentage is almost 15% higher in tournaments than in ladder play. That is an astounding differential for a game like euchre. Now I'm of the opinion you still have to win those games, but perhaps tourny records might function like the imbalanced schedule in baseball or college football. Not all wins are created equal. Looking at it this way, I'm thinking _95crownvic may have to be moved from the "in" category back to the proverbial bubble.

Some other records adjusted for tourny participation:

donotgothere14: Tourny Record (3376-1921)(63%); non-Tourny Record (4950-5235)(48.6%)

dlf416: Tourny Record (263-131)(66%); non-Tourny Record (56.39%)
kenmorefield: Tourny Record (180-77)(70.03%); non-Tourny Record (55.57%)

The interesting thing here is that dlf416, like most players, has a 10-15% higher percentage in tournies than in regular games, but the relatively few number of tournies played in means that number doesn't skew his/her overall percentage that much. (Overall: 56.87; non-tourny 56.39)

On the other hand, sgtpaw2003 has an 0-2 tourny record, making his/her overall 54.6 winning percentage all the more impressive. Hunsbun and yoo_flung_poo_2/Cinbad have remarkably similar 52.6% wins in non-tourny games. Enough to keep their names in the conversation but perhaps not the slam dunks they originally appeared. Moderndaycowboy02 has a 63% tournament winning record but only a 51.5% ratio in straight ladder games. Pimp_mama_frog has a 60% win reocrd over roughly 1800 tourny games, and a 52.98 percentage in straight ladder games that is suddenly looking both better and worse than it might at first glance. (I will point out that this player has a high% of total games played against a smaller pool of opponents, a consideration that may or may not affect one's opinion of many prolific players, who, by nature, play each other more than they play the people who drop by from time to time.)

One could, I suppose, make the counter argument that one can cherry pick one's opponents in non-tourny play whereas it is harder to duck certain players that give you fits in a tourny format. I'm reluctant to say tournies don't matter at all. Maybe it's like tennis where a player might be phenomenal on clay but only average on other surfaces. I guess that is one reason why arguments like this can never be totally by the numbers.

Revised List (in no particular order): tdeem1, klykey2, taco46, breebrat2005, jacmajor, kenmorefield, thorne_4845, sgtpaw2003, yoo_flung_poo_2/Cinbad

Deserve Consideration/On the Cusp: hunsbun, dlf416 (really only reason not in is because I don't know this player and/or don't recall seeing him/her play).

Others? You tell me.

No comments: