2020-21 Bracketology

[shock-it-to-me]

Shox getting national headlines. getting harder for the committee to leave them out.

https://www.usatoday.com/story/sport...le/4612399001/

Uuuum, we didn’t win a “share” of the conference title.

[Kel Varnsen]

Losing "pretty" is more valuable than winning "ugly". Humiliating weak teams is more valuable than winning close games over better competition. That really looks like a formula to reward P5 conferences.

It seems the method of evaluating teams now includes something like this:

"I know they lost a lot of games, but look at how well they played when they lost those games. They must be rewarded for how well they did when they lost those games".

I didn't see KenPom's predictor for the Shox at the start of the season, but I saw others, and the Shox were predicted to have a losing record. I'm guessing the Shox consistently beat KenPom's predictors by winning games he predicted we would lose. Yet somehow the Shox did not get any bump in rankings for beating the predictors. Beating the predictors doesn't seem to count if an argument can be made that a team wasn't efficient in beating the predictors and winning the game.

Everything seems to be set up to actually reward and justify "quality losses". "Quality losses" used to be a joke. Now it seems to be embedded in the evaluation methods. This is insane.

Here's why: we narrowly won games we were supposed to narrowly use, and we narrowly won games we were supposed to comfortably win (ORU, Tulane, Temple, and UCF at home come to mind). Games like Ole Miss, Houston, and Tulsa where we performed above expectations were balanced out by the games I mentioned earlier.

KenPom's ratings and other similar ratings systems reward efficiency...when a team performs above their expected efficiency, they move up in the ratings and vice versa. Not difficult to understand.

The thing I'm frustrated with is that the NET is in the hands of a selection committee that's very imperfect and subjective in how they apply reasoning. No one knows what they're looking for (best teams, best resume in terms of W-L, what is it?), and it causes a lot of confusion and anger on the parts of teams and fanbases.

[wichshock65]

Here's why: we narrowly won games we were supposed to narrowly use, and we narrowly won games we were supposed to comfortably win (ORU, Tulane, Temple, and UCF at home come to mind). Games like Ole Miss, Houston, and Tulsa where we performed above expectations were balanced out by the games I mentioned earlier.

KenPom's ratings and other similar ratings systems reward efficiency...when a team performs above their expected efficiency, they move up in the ratings and vice versa. Not difficult to understand.

The thing I'm frustrated with is that the NET is in the hands of a selection committee that's very imperfect and subjective in how they apply reasoning. No one knows what they're looking for (best teams, best resume in terms of W-L, what is it?), and it causes a lot of confusion and anger on the parts of teams and fanbases.

Did we also start off much lower in the metrics due to our turnover of players? Almost seem we were assumed to have a terrible year and when we haven't, it was impossible to get out of the hole we started in.

[Kel Varnsen]

Did we also start off much lower in the metrics due to our turnover of players? Almost seem we were assumed to have a terrible year and when we haven't, it was impossible to get out of the hole we started in.

The preseason ratings for KenPom and other metrics like it drop off about 15-20 games in the season. So our rating now has nothing to do with where we were at to start the season.

[Shock Top]

So Penn State is expected to lose bad every game, but loses close instead.

So I guess that means they are actually really good.

Seems that’s how the metric are working for them.

[wichshock65]

The preseason ratings for KenPom and other metrics like it drop off about 15-20 games in the season. So our rating now has nothing to do with where we were at to start the season.

We've only played 19 games.

[Kel Varnsen]

We've only played 19 games.

True. So even if they haven’t dropped off completely, they’re having a negligible effect on our rating at this point.

I will also say that KenPom’s predictive ratings by themselves should not be used to select or seed teams in the tournament. That should be left to descriptive metrics like WAB or SOR, which can be calculated using predictive metrics like KenPom. NET is a bastardized amalgamation of the two types.

[TrackSuitAndTie]

WSU is only ranked 78 according to them. So not any friendlier to the Shox.

WSU seems to be really punished for a lot of their game being close. If a few of them were by a couple more baskets, then this might not even be a topic of conversation.

I did notice that. Ideally a metric that is friendlier to us would follow the lead of Haslametrics and stop counting when the game is decided statistically haha.

[wichshock65]

True. So even if they haven’t dropped off completely, they’re having a negligible effect on our rating at this point.

I will also say that KenPom’s predictive ratings by themselves should not be used to select or seed teams in the tournament. That should be left to descriptive metrics like WAB or SOR, which can be calculated using predictive metrics like KenPom. NET is a bastardized amalgamation of the two types.

The metrics this year just seem to be a steaming pile of ****. No way I believe that a team can lose 2/3 of their games and still be ranked 10+ higher. That doesn't past the eye test. Should common sense be a factor?

[Kel Varnsen]

The metrics this year just seem to be a steaming pile of ****. No way I believe that a team can lose 2/3 of their games and still be ranked 10+ higher. That doesn't past the eye test. Should common sense be a factor?

I assume you're referring to Kentucky or Northwestern, in which case, yes that is possible. They have both played very difficult schedules, and while they lost many of those games, it's about efficiency relative to the teams they've played, not resume, W-L, or the "eye test" (in quotations because it's super subjective). Neither team will be making the tournament, and likely not even the NIT to be honest.

You can be a good team and lose a lot of games if you play a really difficult schedule. And you shouldn't make the tournament.

[ShockingButTrue]

Shox getting national headlines. getting harder for the committee to leave them out.

https://www.usatoday.com/story/sport...le/4612399001/

I'm glad a there's a rational mind out there acknowledging the strength of WSU's schedule. Who really believes the selection committee doesn't take into account RPI, in any shape or form? They damn well do when it's convenient.

[shock-it-to-me]

Is it just me, or does it seem like 90% of the “bubble” games in the last two weeks have broke are way? Destiny!

[ShockerFever]

Is it just me, or does it seem like 90% of the “bubble” games in the last two weeks have broke are way? Destiny!

And yet we continue to drop in the rankings.

[shock-it-to-me]

And yet we continue to drop in the rankings.

We’ve dropped in the net but moved up slots in bracket positions. Not very much admittedly

[Aargh]

Let me see if I've got this right. I don't really know, but it seems like it would work like this.

Every team starts with an "assumed" AdjO and AdjD Let me use UNC and Duke as examples. Neither team is very good this year, but pre-season they would have been expected to be better, so the AdjO and AdjD for both is inflated before they play a game. Then they play each other. Both get their AdjD and AdjO adjusted favorably because of the metrics of the opponent. After they play each other, their metrics are even more inflated because they were already inflated. Each team's overrated offense does better than predicted because of the opponents overrated defense. Same thing works for the defensive metric.

If they have other opponents with inflated pre-season metrics, that process will be repeated and the starting value will remain in the formula. It seems the eventual outcome can be greatly affected by the starting assumption, ESPECIALLY in a shortened season.