Hoops Luck

Mar 1, 2022

You have currently viewed 1 story this month.

IlliniBoard now offers two free stories per month, for more please subscribe.

Everybody hates that word: "luck." I say "Wisconsin has been incredibly lucky", and then you tell me they're incredibly clutch, and then I say "that may be but here are the stats" and then some Badger fan will throw a wheel of swiss at me.

It's not an all-or-nothing thing. Teams can be lucky and good. Teams can be unlucky and bad. Some team that has a great record and crazy luck stats is still a good team - they've just been boosted a bit by getting lucky. Some team that has an awful record and "unlucky" stats isn't bad simply because of their "luck" number. Fans defending a coach will say "yeah, but their luck factor was XX so they would this was a good team that just got really unlucky." To me, these stats can't be used like that. They just inform you of what might not be sustainable.

The reason I'm writing about this today: I feel like Purdue beats Wisconsin tonight by double digits. Much of that is the hopeful Illini fan within - I want a shot at sharing the title. But some of that is Wisconsin continuing to walk the tightrope without a net and surviving. At some point, they have to fall.

Maybe the best place to start is to go back to the football "luck" I reference a lot. The stat I like to reference: 2nd Order Wins. You can go find other articles I've written about it if you want to know more (or just go read what Bill Connelly has written about his stat), but here's the basics. Take each game. As Connelly says, "toss all the statistics up in the air" from a certain game and have them land in a random order. Compare them to all other games with similar statistics and teams that generally do X, Y, and Z win these games, say, 74% of the time. Add up all of those percentages and you see how many games you should have won. It's not always luck that changes it, but, you get the point.

As I've always stated, a 99-yard fumble return - one play - can turn a fairly comfortable 34-21 win into a 28-27 loss. You were still the better team. You went up and down the field with ease and forced them to punt a lot. But one play where your tailback wasn't careful with the ball at the one when you were driving to make it a 13-point lead is suddenly returned 99 yards and your offense, despite failing all game, didn't even need to have the ball for you to win it late. (And it doesn't have to happen late - it can happen at any time in the game.)

It's still a great defensive play. Coaching the guy to identify, scoop, and score won you the game. Forced fumbles are a skill. You just can't count on them like you can count on a solid offense to consistently march down the field and punch it in. In the long run, teams that can move and stop the ball almost always overtake the teams that can't. You might get lucky for one game (cough Indiana beating Penn State in 2020 despite being out-gained 488-211 cough), but eventually - probably the very next season (cough Indiana football 2021 cough) - it catches up to you.

I reference all of that to give you the mindset. It's usually not "oh God they're only winning because of luck." Everyone is typically a little bit lucky or a little bit unlucky in a given season. It's just information that informs us (and informs Vegas) about a certain team a certain year.

With basketball, I think this can be fairly simple. When you see a high major team with a really good record and really long odds, there's a good chance their luck numbers are really bad (or good, as it were).

Let's take Providence and Wisconsin. Right now, Providence is 24-3, they won the Big East, and they're ranked 9th. Wisconsin is 23-5, they can win the Big Ten outright, and they're ranked 10th. So when you go to betting sites and look at the odds for Providence and Wisconsin to win a national title, why are the odds so low? I think it's simple: almost everyone knows they're not as good as their record.

On my train back from Ann Arbor yesterday, I looked up the national title odds (and I restricted my search to articles that had come out the previous 24 hours). That landed me on this article about how the top six teams losing on Saturday might have affected national title odds. Here's the national title odds from that article for Big Ten teams (this is as-of Monday morning):

Purdue +1000
Illinois +2500
Michigan State +5000
Ohio State +6000
Michigan +8000
Wisconsin +9000

Wisconsin is 23-5 and Michigan is 15-12 and right now you can get better Wisconsin odds? Well, on that site you can.

Now, a lot of these "who can win the title" lines have a whole bunch of "goad the fans into making a bet that will never cash" in them. Vegas is fairly confident in which teams are a statistical threat to win the title and so they place odds accordingly and then maybe have a little fun further down the list. I'm not trying to make the point that Wisconsin is the 6th-best Big Ten team.

But still, I think it's noteworthy that the two high major teams with the highest "luck" rating (Wisconsin and Providence) are currently 24th and "not listed" on that site. You would expect it to hold that if Providence and Wisconsin are 9th and 10th in the polls they should have the 9th and 10th-best odds of winning it all, but that's not the case. On FanDuel Providence has the 19th-best odds and Wisconsin is not listed. And on this site, offering odds on simply making the Final Four, Purdue is 4th at +220, Illinois is 13th at +700, Ohio State is 16th at +1050, and Wisconsin is 20th at +1400. Providence is on this one as well - 22nd at +1600.

(I'm linking those not as part of some sponsorship thing so that you'll go place a bet or something - I just need to link to these so that you can see that I'm not making this up and that Vegas doesn't think either team has much of a chance in the Tournament at all.)

The 9th-ranked team and the 10th-ranked team are 22nd and 20th in odds on just making it to the Final Four? Why?

Per KenPom's list, luckiest teams as of this morning (out of 358 Division I teams):

1. Providence (+.197)
2. UNC-Wilmington (+.178)
3. Samford (+.172)
4. Texas State (+.157)
5. Wisconsin (+.150)

(Illinois is 132nd at +.024. And very much worth noting, Iowa is 310th at -.062, one of the unluckier teams in the country.)

What is KenPom using to put together this "luck" statistic? I'll just cut and paste his words:

The easiest one to understand is Luck, which is the deviation in winning percentage between a team's actual record and their expected record using the correlated gaussian method. The luck factor has nothing to do with the rating calculation, but a team that is very lucky (positive numbers) will tend to be rated lower by my system than their record would suggest.

Wisconsin's current winning percentage is .821. Their expected winning percentage, based on their season so far, is .671 (that's the difference of +.150 you see there). So this is a 23-5 Wisconsin team with an "expected" record of 18.7 wins and 9.3 losses. As Ken Pomeroy notes there, that's not why his ratings have Wisconsin at #27 when they're 10th in the AP Poll. He's not lowering them because of a +.150 in "luck". It's just that teams that are crazy high on the luck scale are going to tend to be rated lower because their statistics are those of a team (in this case) that would be just outside the top-25.

I always place all of this in the same "toss the stats up in the air" framework that I reference above when talking about football and 2nd Order Wins. Nobody wins every close game. And nobody loses every close game. If this team over here wins 8 games by 10+ and loses two games by 1, and this other team over here wins eight one-possession games and loses the other two by 20+, both teams are 8-2 but the two teams are not equal. One should be 9-1 or even 10-0 and the other one is very fortunate they went 8-2 and not 4-6.

Obviously there's more factors to all of this. I cite that Wisconsin only beat Nicholls State by three all the time, but they played that game without Johnny Davis. There are many other factors that might play into why a team is struggling to put away opponents while other teams are dominating. But in a big broad picture, the baseball team which won the weekend series by winning 2-1 Friday night, losing 15-2 on Saturday, and winning 3-2 on Sunday, can't be counted on the same as the team that won the weekend series by winning 11-2, losing 5-4, and then winning 9-0. Eventually, your 2 runs are gonna match up with someone else's 5 run game, not their 1 run game.

I'm butchering all of this but I think you get the point. When I ran the margin of victory in Big Ten games last week between the three leaders, here were the numbers:

The point of that: what Wisconsin is doing here is unsustainable. Historically, statistically unsustainable. Yes, when you zoom in, maybe Johnny Davis is hitting clutch shot after clutch shot and their fans can claim "this team just finds a way to win". But most every "this team just finds a way to win" eventually fails to find ways to win. The numbers catch up.

That doesn't mean we should decide conference titles on statistics. I'm not saying "if Wisconsin wins tonight and Sunday, it's not a real Big Ten title". There are baseball teams that win a 5-game series despite being outscored 31-12, and maybe that just means they matched up their pitching in a way that could win three games. I'm not suggesting that the team that scores 31 should be declared the winner of the series. You play to win the game.

But if we're trying to inform ourselves on what might happen in the future, that team which was outscored 31-12 in the first round of the MLB playoffs might be in big trouble the next round. Especially if they've been winning games that same way all season. They probably don't have the bats. It's a seven game series now and it's really hard to arrange your pitching in a certain way to win four of those seven. There's probably a tidal wave coming.

For Wisconsin and Providence, that's kinda what the numbers are saying here. I've been telling everyone all day that I think Purdue wins by 10+ because this feels like a "oh, right, Purdue is, like, WAY better than Wisconsin" game. If 12-19 Illinois State almost beat Wisconsin in Madison, and if Nicholls State almost beat Wisconsin in Madison, and if Indiana almost beat Wisconsin in Madison, and if Minnesota almost beat Wisconsin in Madison, and if Penn State almost beat Wisconsin in Madison, then at some point it feels like the luck runs out and a team like Purdue really puts it to Wisconsin in Madison.

Or maybe Wisconsin pulls it out. If so, congrats to them on at least a shared title. You play to win the games and they'd have won the games (even though they'd have gotten a little bit lucky in that they avoided second matchup with the University of Illinois Roadkills).

I said "at least a shared title" there because they'd still have to beat 8-21 Nebraska at home on Sunday to win the title outright. Surely ol' Pythagoras wouldn't catch up with them on the final Sunday and nip them at the finish line, right?

He might.


IlliniJoe81 on March 1, 2022 @ 02:29 PM

Unfortunately, "luck" = Johnny Davis. I've watched the end of a lot of Wisconsin games hoping they would lose and that guy just wins it. Winning these close games is a skill that Wisky has because they have him.

That being said, I agree that if Purdue wins they probably win by 10+. I just can't see the best Purdue team in its history missing the conference title by 2 games.

I'm seeing a 3-way tie for the title, hearkening back to 2002. (I thought for a while that Ohio State might make it 4 teams.) It might take a 10-0 Trent Frazier run vs. Iowa in the final two minutes to get us that last win but it seems like it should happen, just like it felt like we had to beat Michigan on Sunday.

That last Nebraska result was an eye-opener. If it is pro vs pro on Sunday, maybe Nebraska's pro turns out to be a difference-maker?

Anything can happen. Penn State is a hot mess but Iowa is really firing on offense right now. They will be a tough out just like in the Garza days.

It's March and I love it.

Robert on March 1, 2022 @ 03:53 PM

One other note: the unluckiest Illinois team: 2018 (Underwood's first team). 343rd out of 351 Division I teams that year with a -.093 in the luck column. A 14-18 team that really should have been 17-15.

Luckiest? 2004. 47th with a +.055. So a 26-7 team that probably should have been 24-9.

patbasu@gmail.com on March 1, 2022 @ 07:16 PM

Great article- luckiest team in a long time. If both teams play well (and reffing isn’t atrocious) Purdue wins by double digits.

mrmill on March 1, 2022 @ 10:10 PM

Lucky as hell

IlliniJoe81 on March 1, 2022 @ 10:18 PM

Two bank shots from the arc to win the Big Ten. LUCK.

The Olaf Rules on March 2, 2022 @ 12:16 AM

Let them enjoy their BT title; odds suggest they won’t see the second weekend of the NCAA tourney.

Giovantischixstripz on March 2, 2022 @ 08:07 AM

Nebraska peaking at the right time, 2 game winning streak! Cmon Corn Skins!

JJNorthside on March 2, 2022 @ 09:00 AM

Comes back to that saying, you'd rather be lucky than good. Credit Wisconsin for never really putting themselves out of games. That style of basketball, the "let the other team beat themselves" works wonders in this situation.

AGig21 on March 2, 2022 @ 10:25 AM

Big Ten BB Year end review show will be entitled: "Give Me Two Bank Shots and a Conference Title Chaser..."

1701phiillini on March 2, 2022 @ 11:02 AM

A bit of an “Illi noisy” take. If we do not lose to OSU or Purdue at home, another team’s luck doesn’t enter the equation. We had the opportunity and didn’t get it done. Play harder and better.

Many of us complained about Michigan and the BIG championship last year all the way through getting spanked by Loyola in Round 2 while Michigan went to s16.

Let’s get off the collective “but fairness” drumbeat. It’s not the where Coach wants his player’s heads so let’s drop it and get behind them.

ATOillini on March 2, 2022 @ 03:54 PM

A good team, but 12 wins of 5 points or less is telling. Generally I’d say it points out there are really no great teams in the Big 10. Tate had an excellent article before the Michigan pointing out our own issues.

Still think we have the potential to play with anybody, so I’m hoping we get our “luck” come March Madness.

ATOillini on March 2, 2022 @ 04:38 PM

Well, I’m going to comment again. Sent a somewhat derogatory message to a friend of mine before I even read this article stating to him how lucky the Badgers were. It quickly led to him propping an even up bet that they make it to Eilte 8. I never bet on anything, but couldn’t resist. Some local charity is going to get a few extra dollars for one of us per terms of the bet.

orangejulius on March 2, 2022 @ 09:34 PM

After last night they are #1 in luck rating per teamrankings.com and it isn't even close.

Big ten basketball season is short enough that a team can avoid regression.

Our luck rating ? 206.

Speak Your Mind

Please login or register to post comments on the IlliniBoard.

Post Preview