That was quite uninformative and doesn't show any relevant information toward this specific event.
What we did was look at the historical records of just the 63 players entered in this specific event.
The more data that you include, in this case going back 2 years and 82k games, it's absolutely going to average out to a lower number.
This is key. Lots of data makes things average out, yes. But it doesn't wipe out the signal; it reveals the signal if there is one.
The point that previous posts had made, is that for this specific tournament, the winner shot significantly higher than what their average rating is, i.e. they played beyond their average *when it counted*, which is the whole point.
This is actually what we would expect to see for the winner of almost any tournament, because of a selection effect.
Even if everybody is rated perfectly and nobody is cheating you'd still expect something like this.
--On this day, most play their normal B game;
some find their A game, and others play their C game
-- for most players good and bad rolls tend to even out for the day, but
some have net good rolls and others net bad rolls.
--some have a path through good-day/good-rolls opponents;
others face bad-day/bad-rolls opponents.
When you select by success (look at tournament winner), you're preferentially selecting from people with multiple of the bolded things going on. That's essentially a survivor-bias effect
What you are attempting to conclude is that by looking at the bigger picture, everyone is playing within their means and this is all under the typical bell curve. It's certainly the ideal way to look at things, and with a big enough data sample, it's extremely easy to justify your statements. However, I'm willing to bet the numbers tell a different tale when you look at the winners for events where big money is on the line. Will some play to their rating? Of course. But will a few far exceed what they should be capable of? Definitely. These are the people in question.
This is not a random sampling of players; It is is a group of 63 players who are unusually incentivized to perform better "when it counts." And what we were able to do is interrogate their histories looking for any tendency to "perform better when it counts." Did they perform worse in league or in small don't-matter events than they did in big high-dollar events? The answer was no--we didn't see any evidence of that.
[...] Those with ill-intent will game this system as they do with all others. This all has no bearing on me as I don't play in any fargo events, but it's certainly fun to watch from the sidelines!
While sure, there are people who try to game the system, most of the things you're "watching from the sidelines" are false alarms. But we have no way to later update that what we heard was a false alarm. So we nurture a perception that fires are blazing everywhere.
It takes a dive into the data to answer the question...
how many houses are burning down--really?. Our job is to use that information to recalibrate our perception.
www.facebook.com
