[POOL] Argonne Pool League Week #10

[John Valdes]

If only there were some company that took historical performance data,
ran it through some sort of algorithm, perhaps involving a computing
machine that learns somehow, and tries to predict future performance
outcomes. :)

I'll pass the feedback on to Dean, thanks.

John

On Fri, Nov 22, 2019 at 05:10:49PM -0500, Paul wrote:
> I did a quick analysis after Dean sent the proposed change out. Of my last three matchups:
> Tim – 72,
> Frankie – 108,
> Dennis – 89,
> Only the matchup with Frankie would have changed. I played him in a 8:2 which would become 7:2 with the new table. In summary, what Paul said. 
> Paul Lucido
> 
> ________________________________________
> From: mcs-pool [mcs-pool-bounces at mcs.anl.gov] On Behalf Of Hovland, Paul [hovland at mcs.anl.gov]
> Sent: Friday, November 22, 2019 9:36 AM
> To: MCS Pool League - Moe
> Subject: Re: [POOL] Argonne Pool League Week #10
> 
> Captain John / Team Moe:
> 
> FWIW, I’m not a fan of completely eliminating 3:1 in the 10-39 bracket.  A 3:2 matchup for a 39 handicap versus a 0 handicap seems unfair.  Maybe 34-39 should be 3:1.
> 
> I’m also not a fan of the shift in the 100+ bracket. I’d like to see some empirical evidence that the X:1 matchups have been won disproportionately by the lower handicap players. I don’t think this is true, at least for the 25-50 handicap players.  I’d be okay with introducing a new 130+ or 140+ bracket and using the proposed 100+ bracket there.  This would remedy the 150 v 66 and 163 v 76 examples that Dean cites without adversely affecting 122 v 28. Introducing a 130+ or 140+ bracket would also make it easier to introduce some variant of Dean’s proposed rule to handle the 150 v 0 matchups.
> 
> In general, I’m worried that our handicap table, which was originally based on a lot of careful analysis (documented in the rule book), has undergone a long sequence of ad hoc adjustments, each intended to remedy one perceived problem but potentially introducing new ones. It might be time to build a new handicap table from scratch. I don’t know whether Ron and the other developers of the original handicap table would be willing to repeat their analysis, based on the current handicap rules and possibly supported by years of empirical data (unfortunately, if a lower handicap player won a 7:1 matchup, we have no way of knowing what would have happened in an 8:2 matchup, but if the higher handicap players are winning at least 50% of such matchups, it wouldn’t matter).
> 
> Paul Hovland (don’t blame Lucido for this rant)