Mike Beuoy

actuary | creator of | co-creator of |
you can also find me at bsky under the same handle

[Mike Beuoy @inpredict]

I'm so excited for this launch, and incredibly proud of the work @cwetzel31 and @basketballvalue have put into this.

I'm really looking forward to how we can contribute to the sport of volleyball.

[Evollve @evollve]

We are thrilled to announce the launch of Evollve!

Our goal at Evollve is to provide the volleyball community with consistent, reliable, and easily accessible data and analysis.

Check it out at https://t.co/TdHfRxmV83!

[Mike Beuoy @inpredict]

background: https://t.co/YFW636MC20

[Mike Beuoy @inpredict]

Midterm Model Melee update:

Consensus Senate Forecast has moved away from the Deomcrats - this leads to a drop in bankroll/credibility for the Silver Bulletin model since it is highest on the Democrats relative to the consensus.

[Mike Beuoy @inpredict]

A good illustration of where models and intuition disagree.

I think a lot of people, including coaches, think in terms of “scores behind” when it’s late in the game. So a 12 point deficit doesn’t feel any better than a 9 point deficit - it’s still “two scores”.

[Seth Walder @SethWalder]

Pretty funny after this week's #discourse but it actually happened: Kyle Shanahan went for it on 4th down when the model preferred kick!

[Mike Beuoy @inpredict]

@bigfishlurking k

[Mike Beuoy @inpredict]

I wouldn't necessarily call it variance, but this tracks to the data I've pulled.

A team that is up 20 at the half underperforms by about 5 points in the 2nd half.

[Bobby Krivitsky @BobbyKrivitsky]

Joe Mazzulla says the Celtics analytics staff studied situations where teams led by at least 20 over the last decade.

He says that when you’re up by 20 in today’s game, not only is growing that to 30 not a realistic expectation, but due to variance, you’re really up by 15.

It will most likely shrink to 6 or 7. It comes down to how you then handle that situation.

[Mike Beuoy @inpredict]

@WeatherMaker007 Yes. No plays excluded.

[Mike Beuoy @inpredict]

@ryanwold The numbers are for the team with possession - I did that so each data point only got counted once.

[Mike Beuoy @inpredict]

@jasontconnor Mean reversion would be zeroes.

[Mike Beuoy @inpredict]

@bhboehlert I don't think so. This is all forward looking. Mean reversion would be a chart of 0's across the board.

[Mike Beuoy @inpredict]

@bhboehlert I don't think so. This is all forward looking. Mean reversion would be a chart of 0's across the board.

[Mike Beuoy @inpredict]

@BradChilmeran i think it's both

[Mike Beuoy @inpredict]

Communicating uncertainty in your probability estimates

[Blake Murphy @BlakeMurphyODC]

Net rating stabilizes around 1 game if you really want it to

[Mike Beuoy @inpredict]

@skepticalsports Potentially, although I think the average coach still under-optimizes under that standard. I think plenty of fans want their team to go for it.

But adding a large multiplier effect for win probability lost on failed 4th down conversions would probably capture fan engagement.

[Mike Beuoy @inpredict]

Win probability is not a bankroll you need to manage with a calibrated fractional Kelly strategy.

You're not going to parlay that 0.10 of win probability into a long term optimal growth rate.

There's no benefit in avoiding ruin if there is an option with a higher *expected* WP.

[Mike Beuoy @inpredict]

4th and 6? Too risky to go for it. Let’s play it safe and send out my college kicker for a 55 yard attempt.

[Mike Beuoy @inpredict]

@benbbaldwin The Rams 2-3 record is like the peacock’s tail. An encumbrance that functions as a signal of underlying strength.

[Mike Beuoy @inpredict]

@peanut_bettor Not relevant to the Commanders situation, but just a general comment regarding coaching tendencies.

Kicking as soon as you're in "range" doesn't maximize win probability. It maximizes the probability you can blame somebody else for losing.

[Mike Beuoy @inpredict]

Some coaching decisions can be explained by assuming they view losing a game due to a missed field goal to be preferable to losing a game on a turnover or a failed 4th down conversion.

[Lord Commander 999 @HailMaryland]

"As soon as you are firmly in field goal range you send the kicker out, it's as simple as that!"-Mike Tomlin

[Mike Beuoy @inpredict]

New Substack post: Midterm Model Melee

Pitting 3 forecasts of the Congressional Midterms against each other in a daily betting/prediction contest.

https://t.co/5OJDN4ZXom

[Mike Beuoy @inpredict]

@alejandroggo Yeah I think that’s probably right. But it also means there’s just fewer games.

Part of it is also just G5 of the 2016 Finals stuffing the ballot box.

[Mike Beuoy @inpredict]

Putting these two charts side by side, what jumps out is that the top shot in the NBA (+28% title probability), would rank just #6 on the WNBA list.

And that the 2016 Finals shows up on both of these charts.