Home/Away Splits In March

The comments of the "Original Bracketology" post are really excellent. David pointed out what looks like a trend in the data from teamrankings.com. They have separate home and road ratings for each team, dating back to 98-99. He noticed that teams with good home rankings, but poor road marks, tend to fail pretty spectacularly in March:

"I looked for all teams in that span that were:
A) ranked 20th or better at home
B) ranked 51st or worse on the road
C) seeded 1 through 6

Year #Seed Team (Home/Away) - Result
----------------------------------------------------------
2006 #3 Iowa (2/53) - 1st round upset
2005 #6 LSU (15/51) - 1st round upset
2003 #6 Missouri (17/57) - 2nd round loss to #3 seed
2002 #6 Texas Tech (17/51) - 1st round upset
2001 #5 Virginia (4/56) - 1st round upset"

This is obviously not a very large sample, but it does look like there's something here.

I collected data on the top six seeds in each region from the '99-'06 tournaments (excluding '01, since the teamrankings.com data is incomplete). On the right are the expected wins for each seed. This gives us a baseline to fairly assess the wide range of teams that we are looking at.

Using this, we can see how each team did compared to their expectation. For example, if a #1 seed wins two games, they're -1.04. But if a 6 seed reaches the Sweet Sixteen, they're +0.75.

Next, taking David's idea, I took the teams with the biggest differentials between their home and away rankings. There are 168 teams in the sample, so let's look at the top and bottom 10%. Here are the teams with great home ranks that were awful on the road:


This is pretty amazing. 15 of these 17 teams won fewer games than expected*. The average "vs Exp W" is -0.50, which is remarkable. Let's look at the other end of the spectrum; teams with weak home ratings, that were good on the road:


It's not as extreme, but we do see the opposite trend. These teams win an average of 0.24 more games than we would expect.

This trend is both anecdotally and statistically significant. I did a regression using these 168 teams. The independent variables were their home and away ranks, and with the dependent variable (what we're trying to predict) being tournament wins. Since a lower rank should lead to more wins, we would expect both estimates to be negative. Here are the results:

The minuscule P-Values show that both variables are statistically significant.

The absolute value of "Away" is 58% larger than that of "Home". Clearly, a team's road performance is a much better predictor of tournament success than what it does on it's home floor.

What does this all mean? Well, let's apply it to this year. Since the data only considers the top six seeds, we'll stick to that- conveniently, that's what I covered in the Bracketology post. So, of those 24 teams, who has the most dramatic home/road splits?

I think sufficient attention has been paid to the first team on this list. I'm not at all surprised that Notre Dame has a significant split- they were 9-0 at home in the Big East, but only 5-4 away from South Bend.

Xavier was a little surprising, but understandable. They are 16-1 at home, with their only loss coming to Tennessee. Their road record is 7-4- none of these opponents have been particularly good, and they lost to the best one, Arizona St., by 22 back in December.

It should be noted that these two teams have the second and third largest differentials, but they're not huge. So I'd say it's a slightly negative indicator, but nothing to be too worried about. This does not apply to Vanderbilt, obviously.

The second half of the list is interesting as well. Washington St. has been tremendous on the road, with wins over Baylor, Gonzaga, USC, Arizona St., Oregon, Cal, and Washington. To this point, they've been held back by their disappointing 5-4 home record in the Pac-10. This is a dangerous team.

I am surprised that Drake appears here. I suppose road wins against Butler, Illinois St., Creighton, and Bradley help. I am still not sold on them, but maybe my skepticism is misplaced.

Update: I forgot to include Michigan St., because I didn't have them as a top 6 seed. In reality, they're probably a 6 right now. They are 3rd in the home rankings, and 59th away; they'd be right behind Vandy.

*Because of the weird distribution of wins (each year, one team wins six, one wins five, two win four, four win three, and so on), 65% of teams are below their expected wins. In our example, 88% of the teams in the first table (bad road teams) won less games than expected, while for the second table (good road teams) only 47% won less than expected. This confirms what is displayed in the average (road success is a better indicator than home games), but it's important to have a relevant baseline.

Finding Value In Futures Odds

Last week, in the MVC and CAA conference tournament previews, I suggested that both league champs, Drake and VCU, had favorable odds. Drake ended up winning their tournament, while VCU lost to William & Mary in the semis. As it turns out, VCU would have been a much smarter wager. Let me explain.

The Greek had VCU at +175. In their first game against Towson, they were favored by 14; their money line was about -1200. Today, against W&M, they were 12 points favorites, with a money line of -800. If they had won, they would've faced George Mason; that game probably would've been around a pick'em (-110). So, if they had won out, and you had bet $10 on their money line in the first game, and carried that money over throughout the tournament, you would've ended up with $23.67; their implied odds for winning the tournament, from the lines of the actual games, were +137*. That makes +175 quite the bargain.

The opposite was true for Drake. Since they reached (and won) the final, we conveniently have the actual lines to all their games, which were 10, 3, and 2.5, respectively (Drake being favored in each). The corresponding money lines for these spreads are -500, -155, and -145, respectively. If you'd started with $10 and continued putting it all down on them for each game, you would've netted a profit of $23.36*. In contrast, putting that $10 down on their odds at the beginning of the tournament (+175) would have only given you a profit of $17.50.

The outcomes for these two teams are irrelevant. By betting on Drake to win, you are doing yourself a disservice. On the other hand, VCU's line of +175 significantly better than the market value determined by the individual game lines. When considering futures like this, one has to not only figure out if the odds are profitable by themselves, but also assess whether that is the best option.

*Obviously, this depends on who each team ends up playing in each round. Drake faced the #8, #4, and #2 seeds, which is the most difficult possible path. VCU went up against the #9 and #5 seeds, and would've faced #3 George Mason in finals. This has a marginal effect, but nothing extreme. It's interesting to note that George Mason was favored by 7 over the #2 seed, NC Wilmington (which, as previously noted, was vastly overrated). So Creighton would have been favored substantially in the championship game over NC Wilmington, likely making their implied odds even lower than +137 (lower in this case meaning closer to EV, +100).

'08 Conference Tournament Odds: MVC

The Greek has posted odds on the MVC tournament, which begins on Thursday.

Those odds are below. I have also used the Pomeroy ratings to determine what % chance each team has of winning each matchup, and thus their odds of winning the whole thing. I then translate that into what their odds should be. This isn't a perfect system, but it gives us a pretty good idea.

For the most part, the "Pomeroy" column will have higher odds than the "Greek" column. This is because the percentages I calculated obviously add up to 100%, while the Greek's odds add up to 138.6%. But occasionally the odds are higher than they should be; those are the profitable bets.

I would advise against betting on Evansville.

It's very surprising to me that the Drake odds are worth betting on. They've gotten a good deal of national attention; it seems like people would bet on them at +150, and even lower than that. I also think most people would see the best team in the conference with a "+" next to their odds and think it's a good line. It's unclear why their line is as high as it is.

SIU is not a bad bet either. The Salukis have had a down year, but they're actually right there with Illinois St. in the tempo-free stats, which shows in how similar their Pomeroy odds are to the Redbirds'.

(I have very little school work over the next two weeks; you should probably be checking this site every 10 minutes or so.)

Blogpoll: Week Fourteen

The full blogpoll is here, my ballot was as follows.

Indiana is screwed. The Sampson mess, the Butch three- not a good day for the Hoosiers. I've thought they were a little overrated for a few weeks now- they're now 9-2 in the B10, but they could easily be 7-4 or even 6-5. But now, if they lose to MSU and Purdue at home in the next week (both at home), it'll all be blamed on the coach. Not that that situation doesn't have any effect, but it's probably not the whole story.

UConn is incredible- tonight makes eight in a row (gotta love Thabeet's line of 4pts, 10 boards, 6 blocks). They were a 4 in Bracketology this week- they may even get bumped up a line on Monday. Their last seven BE games really aren't that tough- they play at USF, Villanova, and Providence, and host West Virginia, but will likely be favored in every game the rest of the way.

Louisville should certainly be higher (although I'm closer than some). It's always tough to know what to do with teams that were missing key players early in the season. Despite having one more loss than Georgetown, they've been the best Big East team in conference games. They have a much more difficult schedule than the Huskies, with games @Pitt & Georgetown, and home games against Notre Dame and Syracuse.

Drake lost in Carbondale tonight, but I won't drop them that far next week. You are allowed to lose once every couple months, I think.

Blogpoll: Week Thirteen

The full blogpoll is here, my ballot was as follows.

I didn't watch any of the game tonight- I was in the library studying Corporate Finance. I see that Lawson didn't play, and the combined 7/34 shooting of Ginyard, Ellington and Green didn't help matters. Scheyer got six rebounds? That's interesting. Looks like Duke's 39-9 advantage from beyond the arch more than made up for UNC's dominance inside. I don't think people will ever stop saying "Duke has no inside presence"; we'll see how far they can advance despite that.

Drake just keeps on winning. They're up to 20 in a row after coming back to win at Indiana St. last night. The Valley is down a little this year, but starting off 6-0 on the road is difficult no matter what conference you play in. They've now got another big road test on their schedule, as they'll visit Butler on February 23 as part of BracketBusters weekend.

As detailed over at RTC, the Gonzaga-St. Mary's game was pretty incredible on Monday night. Pargo and Mills really took over down the stretch, and St. Mary's got pretty much every call in an extremely hostile (and dark) environment. The Gaels are at Santa Clara next Monday, and their next real test after that will come against Kent St. in another BracketBuster game.

Notre Dame was very impressive tonight (not that I saw the game, but they did win by 26) on the road against Seton Hall. They're starting to shed their reputation of a poor road team- they also won @Villanova a couple weeks ago. They host Marquette on Saturday (a game they really should win), and then play @UConn next week- that should be an excellent game.

Purdue makes their first appearance on my ballot this week. They've started off 9-1 in the Big 10, but now their schedule really gets tough. Their next four games are at Wisconsin, vs. MSU, at Northwestern, and at Indiana. You'd have to consider a split of those four games (so, really, beating Northwestern and winning one of the other three) a success. If they do that, they could very realistically finish 14-4 in the Big Ten, which is almost shockingly good.

Checking In

A couple weeks ago I had two posts (linked below) examining the biggest discrepancies between the AP poll and the Pomeroy ratings. Today I'm going to go back and see how these teams have done since. First, those who were...

"Under the Radar"

Marquette
Then: AP #13, Pomeroy #3
Since: 3-3
Now: AP #16, Pomeroy #9

I was very wrong on this one. Marquette added two more ugly road losses to their resume, at the hands of Louisville and Connecticut. It looked like they might be turning around after they finally got a conference road win against Cincinnati on Saturday, but then they turned around and got blown out at home by Louisville tonight. Definitely not a top ten team, and both of their current rankings will fall after tonight's game is factored in.

Xavier
Then: AP #20, Pomeroy #9
Since: 4-1
Now: AP #13, Pomeroy #10

Xavier was definitely underranked at 20, and may still be at 13. They surprisingly lost @Temple by 19, but have been excellent since, beating a shorthanded Dayton team by 26 at home, and winning @UMass by 12. They have three of their next four on the road, so it might be tough to stay this high in the polls, but luckily that doesn't really matter. Lunardi currently has them as a 3 seed in the NCAAs- they will likely end up as a 3 or 4, and are certainly not a team I'd want to face.

Drake
Then: AP #26, Pomeroy #15
Since: 6-0
Now: AP #15, Pomeroy #25

This is interesting. Drake's win streak has reached 18, allowing them to move all the way up to 15th in the AP. Yet they've fallen ten spots on Pomeroy's ratings, since each of their last six wins have come by ten points our less. They also have a very difficult upcoming schedule, with visits to Illinois St., Southern Illinois and Northern Iowa in the next two weeks. I bet they'll lose one or two of those, and fall back under the radar until March. Lunardi currently has them as a 3- I'd be surprised if they stay that high, they'll probably end up at 4 or 5.

Wisconsin
Then: AP #17, Pomeroy #6
Since: 5-1
Now: AP #8, Pomeroy #5

The Badgers are the best team in the Big Ten. I don't see all that much room for debate at this point. Michigan St. has suffered borderline embarrassing losses against Iowa and Penn St. Indiana somehow lost to UConn at home, and then lost @Wisconsin by 13. Meanwhile the Badgers are 8-1, with their only loss coming at a surprisingly tough Purdue team. They do have five of their last nine on the road (although the last one is @Northwestern, which barely counts), but should be able to scure at least a 3 seed in March.

And the teams I was quite critical of:

"They're Not That Good"

Vanderbilt
Then: AP #16, Pomeroy #65
Since: 2-3
Now: AP #23, Pomeroy #75

Vandy is still ranked? Seriously? Let's look at their SEC road games:

1/12 Kentucky; lost 79-73 in OT
1/17 Tennessee; lost 80-60
1/27 Florida; lost 86-64
1/30 Mississippi; lost 74-58

They're 3-4 in a weak SEC because they've beaten South Carolina, LSU, and Auburn at home. I honestly have no idea how they are still 23rd in the AP, and 20th (!) in the ESPN poll.


Dayton
Then: AP #14, Pomeroy #44
Since: 1-4
Now: Zero AP votes, Pomeroy #69

Dayton has been really bad. They've lost to UMass, Xavier, Richmond and Rhode Island by an average of 16 points.

The Flyers do have a good excuse. They've been without freshman Chris Wright (broken ankle) for this entire stretch, and Charles Little (broken foot) missed the Xavier and Richmond games. Little is back now, but Wright won't return until March. Their schedule down the stretch isn't bad- they've gotten all their difficult road games out of the way- but they're down to a 10 seed in the latest Bracketology, and that 25-point win against Pitt seems like a distant memory.

Villanova
Then: AP #25, Pomeroy #73
Since: 2-5
Now: Zero AP votes, Pomeroy #89

This overrated post was pretty solid. Villanova has been the worst of the three- after getting demolished by St. Joe's on Monday night, they've now lost five straight. At 3-6 in the Big East, they won't make the tournament unless they catch fire down the stretch. They were certainly never one of the 25 best teams in the country, that was a product of going 10-1 against a weak (#216 in the country) non-conference schedule.

Blogpoll: Week Eleven

The full blogpoll is here, my ballot was as follows.

This thing is really a mess this week.

I expected to look like an idiot with this if Drake lost to Creighton last night (Creighton was favored by six), but they won in overtime, 68-60. Kyle Whelliston predicted the exact score, which is pretty impressive. It was Drake's first win @Creighton since 2002, and their best win of the year so far. They have a good shot at starting off 11-0 in the MVC before a trip to surprising Illinois St. on February 5.

This ballot does end up looking kind of silly, since Tennessee somehow lost to Kentucky last night. I suppose it is difficult to win on the road when you shoot 7/26 from three and get outrebounded by 11. I still think the Volunteers are the best team in the conference, but maybe the gap is smaller than I thought.

Ole Miss was 15th in the overall blogpoll, but I don't have them ranked. This will continue to be the case until they win a road SEC game. Mississippi St. on Saturday would be a good place to start.

Pretty much the same situation for Vanderbilt. In their two SEC road games they lost to Kentucky in double OT (which was really a terribly played game on both ends, although that loss doesn't look as bad after last night), and lost to Tennessee by 20. They play @UF and @Ole Miss in the next week, so they'll have no shortage of opportunities to prove themselves.

I'm honestly not really sure what inspired me to put Gonzaga so high. Although that win over UConn in Boston could end up looking pretty good. They will probably get blown out at Memphis on Saturday, and that'll be the end of that. St. Mary's is #15 because they handed Drake their only loss, beat Oregon, and have only lost @Texas and @SIU. They are also 4-0 on neutral floors. They should move to 6-0 in the WCC before a home game against Gonzaga a week from Monday.

I have West Virginia 18th, which is the highest of anyone in the blogpoll (they are unranked in the full ballot). Their only losses are in road/neutral games to Tennessee, Oklahoma, Notre Dame, and Louisville, and they've beaten Duquesne, Marquette, and Syracuse at home. They play @Georgetown on Saturday in a game that Pomeroy's numbers actually have them winning, so that'll be interesting. More on the Hoyas later tonight.

Some interesting matchups/lines tonight (for the record, this is not, in any way, shape, or form, betting advice; it's really the opposite of that, as the lines that the books set tend to show us when the perceptions of teams are skewed):

Missouri @ Texas Tech (-1); suprising that TT is only laying a point here, as Mizzou is 1-6 in road/neutral games, while Tech is undefeated at home, including last Wednesday's win over A&M.
Baylor @ Texas A&M (-9); A&M laying a whole lot of points to a good Baylor team that's already beaten South Carolina and Nebraska on the road.
Kansas St. (-3) @ Colorado; really short line here, as K State just won @Oklahoma, and Colorado already has eight losses.
Florida (-1) @ South Carolina; another surprisingly small number
UNC (-7.5) @ Miami, Memphis (-13) @ Tulsa; Carolina and Memphis getting 70% of the action on the road, as usual

Under the Radar

I thought it would be interesting to contrast the latest AP rankings with the current Pomeroy ratings (explained here) to see where the largest discrepancies lie. Today I'll look at teams that haven't gotten as much attention as they deserve; tomorrow, teams that aren't as good as their AP rankings might indicate.

Marquette; AP #13, Pomeroy #3

I was initially surprised to see Marquette this high in Pomeroy's rankings, but now I get it. They don't have any marquee wins, but their resume is impressive in more subtle ways. There are two main characteristics that cause teams to be underrated by the AP poll, and Marquette has both.

1. "Quality" losses

Marquette has lost two games. In November, Duke beat them by four in Hawaii. Then, a week ago, they lost at West Virginia by 15. Neither of these are bad losses; everybody knows about Duke, but West Virginia is very good as well. According to Pomeroy, Marquette has played the 23rd most difficult schedule in the country.

2. Destroying teams

The AP voters mostly look at whether you win or lose, rather than how much you win or lose by. When we're trying to predict a team's future performance, doing this is ignoring some important information. Three of Marquette's victories are good examples of this:

11/20 Oklahoma St. (N), 91-61
1/3 Providence (H), 96-67
1/12 Notre Dame (H), 92-66

Having beaten these three teams, none of which are ranked, isn't particularly impressive. But beating them by an average of 28 points bodes very well for Marquette as they continue Big East play.


Xavier; AP #20, Pomeroy #9

Xavier has lost at Miami (OH) and Arizona St., and at home against Tennessee. The fact that they lost to ASU by 22 is a concern, but the concern is lessened when you look at their final three non-conference games:

12/31 Kansas St. (H), W 103-77
1/3 Virginia (H), W 108-70
1/6 Auburn (A), W 80-57

To me, even more impressive than the average MOV of 29 points was how Vegas dealt with these games. They were favored by 9 against K State, 9.5 against UVA, and 9 @ Auburn. Each of these lines were higher than I would have expected. The books clearly didn't want a lot of money on Xavier in these games, and the results explained why. Dayton has gotten most of the attention, but Xavier is the best team in a very strong A-10.

Drake; AP #26, Pomeroy #15

The Bulldogs lost at St. Mary's in their second game of the season, which is understandable, as St. Mary's is undefeated at home this year. Drake has since run off 13 straight victories, and are getting dangerously close to the Top 25. Their two most impressive non-conference wins came in beating a very good Duquesne team by four, and beating Iowa St. by 35. But it's been their MVC games that have allowed them to rise in Pomeroy's rankings, as they've started off 5-0, winning each game by an average of 11.4 points.


Wisconsin; AP #17, Pomeroy #6

The main reason that Wisconsin is only #17 is they were 40th in the preseason poll. It would be difficult to move up higher than 23 spots while losing twice over the first couple months of the season. But what's important is that those were quality losses; @Duke, and against Marquette at home. Their win @Texas is very impressive; it's also their only win that has come by less than 10 points. They haven't had much trouble in their first three Big Ten games, easily beating Michigan, Iowa and Illinois.