A look back at preseason predictions

[Lennys Tap]

Those are facts. None of them can you get to the conclusion that we would haven't still gone 4-0 against Creighton/Xavier. You can make an educated guess. But that's all it is, a guess. Basketball is not played on paper. We were a significantly better team than Providence last year but we still went 0-2 against them.

Personally, I think we would have gone 3-1 against them if they had been fully healthy. Which probably would have been just as favorable as going 4-0 against them being depleted.

Odds say the most educated guess would be 1-3. Next would be 0-4, then 2-2. Second least educated guess would be 3-1 (6% chance of occurring - .1x.1x.5 +.1x.1x.5 + .5x.5x.1 + .5x.5x.1 = .005 +.005 +.025 +.025 =.06) beating out only 4-0, which would occur approximately 2.5 times per thousand (.1x.1x.5x.5 =.0025).

Certainly anything is possible, but the people who turn facts into probabilities in Vegas for a living would say you have an extremely low probability of being right. That's a fact.

[barfolomew]

Basketball is not played on paper.

Basketball is played on the precursor to paper, depending on the mill.

[TAMU, Knower of Ball]

Odds say the most educated guess would be 1-3. Next would be 0-4, then 2-2. Second least educated guess would be 3-1 (6% chance of occurring - .1x.1x.5 +.1x.1x.5 + .5x.5x.1 + .5x.5x.1 = .005 +.005 +.025 +.025 =.06) beating out only 4-0, which would occur approximately 2.5 times per thousand (.1x.1x.5x.5 =.0025).

Certainly anything is possible, but the people who turn facts into probabilities in Vegas for a living would say you have an extremely low probability of being right. That's a fact.

I have no idea what all those numbers are or where they came from. I don't care about gambling odds. They aren't real stats.

Let's use actual stats.

Mo Watson had a Value Add of 4.87 points per game last season. We beat Creighton by 8 twice.

Trevonn Blueitt had a Value Add of 7.17 points per game. Edmond Sumner had a value add of 6.49 for a total of 13.66. We beat Xavier by 22 the first game.

Edmond Sumner had a value add of 6.49. We beat Xavier by 11 the second time.

Now I am not suggesting that this is proof that we would have gone 4-0 against Creighton and Xavier had they been fully healthy. That is impossible to say. Which has always been the point I was making. We have NO idea what happens if Waston/Sumner don't go down. We can make guesses but that is all they are, guesses.

[jsglow]

Timing in life is everything.   

[Lennys Tap]

I have no idea what all those numbers are or where they came from. I don't care about gambling odds. They aren't real stats.

Just because you don't understand predictive statistics (Pomeroy, Nate Silver, etc.) or simple multiplication doesn't make them "not real".

Let me explain what those numbers are and where they came from. With their full rosters, Xavier and Creighton were approx. 10 point favorites vs MU at home and 1-2 point favorites on the road. This means (accordinding to real statisticians like Pomeroy or Silver)) that MU had roughly a 12% chance to win the 2 road games and a 45% chance to win the 2 home games. I rounded the road games down to 10% and the home games up to 50% for simplicity. So the mathematical chance that MU would have gone 4-0 given those odds are .1 (10%) x .1 (10%) x.5 (50%) x .5 (50%). So the fair value over/under on how many times Marquette would win all 4 of those games if X and Creighton are at full strength is 2.5 times per thousand. Conversely, the fair over/under on MU going 0-4 is .9 X .9 x .5 x .5 (202.5 per thousand)

You can dismiss these as "gambling odds" but I let you wager versus Vegas if you let me wager versus whatever formula you care to use. If your numbers don't jibe with Pomeroy's, Silver's, Caesar's, etc., you'll end up very broke and I'll end up very rich.

[brewcity77]

So the mathematical chance that MU would have gone 4-0 given those odds are .1 (10%) x .1 (10%) x.5 (50%) x .5 (50%). So the fair value over/under on how many times Marquette would win all 4 of those games if X and Creighton are at full strength is 2.5 times per thousand.

That could be the case for any set of games. If we were 60% favorites in four home games, we'd only go 4-0 12% of the time.

I think it's significant that we didn't just win those games, we won them fairly comfortably. Might one or two players have made a huge difference? Sure. Entirely possible. But certainly not a given. No matter what the percentages say.

[Lennys Tap]

That could be the case for any set of games. If we were 60% favorites in four home games, we'd only go 4-0 12% of the time.

I think it's significant that we didn't just win those games, we won them fairly comfortably. Might one or two players have made a huge difference? Sure. Entirely possible. But certainly not a given. No matter what the percentages say.

As to paragraph #1 - 12.96% of the time to be exact, so 13% is more correct. Those are the true (and fair) odds of such an occurrence (given your probabilities) happening.

As to paragraph #2 - meaningless, besides the point information/speculation /opinion that does nothing to alter the indisputable fact that the odds/chances of a team winning 4 straight if the games are pick 'em, pick 'em, +10 and +10 are around 400-1 (2.5 in 1000).

[brewcity77]

It's exactly as meaningless as the speculation of what might have happened had unavailable players played, and the presumption that odds always go to form (I'd point to our odds of beating 'Nova when we were down 17).

The reality is what happened, not what might have happened. If nothing else, it's only worth even discussing for three of the games. Watson was never going to play the season finale with a felony hanging over his head, regardless of his health.

[Jay Bee]

A look back at LNH's Preseason Predictions:

http://latenighthoops.com/revisiting-marquettes-2016-17-preseason-projections/#.WSro_-srLIU

"We said, "A net improvement of 2.5% in eFG% differential means a +5.8% and places Marquette in or around the top 35 of eFG% differential, by far the most important of the four factors. Do that, and they are in business even without improvement in rebounding."

As of Selection Sunday, Marquette's actual eFG% differential was 5.7% (57.6% on offense, 51.9% on defense) and they were selected to the tourney's field."

[Marcus92]

Thanks for posting, Jay Bee. Still love the reference to Katin's rebounding allergy. This season will be tough to predict. JaJuan, Katin and Luke were all among the team's leaders in minutes, possessions and shots. But none of them had complete games.

JaJuan's inside-outside game, rebounding, passing and steals were compromised by inconsistent shooting and turnovers. Katin's offensive prowess was offset by a low free throw rate and the lowest defensive rebounding percentage on the team — behind Markus Howard and tied with Andrew Rowsey. And while Luke had one of the team's highest offensive ratings, he never developed into a strong rebounder.

Returning players Markus Howard, Andrew Rowsey, Haanif Cheatham, Sam Hauser and Matt Heldt provide a strong foundation to build on — with plenty of game experience between them. But the team will need solid contributions from multiple unproven newcomers to make it back to the NCAA tournament.