Posts Tagged projections

Prognostication Corner: Your Headquarters For Misapplied Formulas and 51 Percent Accurate Baseball Picks

Now that Tax Day has come and gone, the only thing left to dread about mid-April may be the ungodly formatting and scheduling of the NBA playoffs. They begin on Saturday, when the first games of four best-of-seven series are to be played.

The most important tournament in professional basketball is often marred in its early rounds by four-game sweeps or similarly short series which lead to first- and second-round games being played during the same time period. We can probably expect this tradition to continue this year, assuming the NBA hasn’t wised up yet — and why would they? — and decided to have a more organized playoff schedule. The probability of multiple rounds being contested at once is particularly high with a stacked Western Conference producing several high-profile first round matchups while a top-heavy Eastern Conference continues to lag behind.

Anyway, it’s pretty late now and I actually have to be up somewhat early to run a couple of errands that I’ve repeatedly put off throughout the week. So I’ll cut to the chase. I’m letting the wacky formula that served me so well in March once again try to carry the day in for the NBA Playoffs. This is a concern for a couple of reasons. First, the college and NBA games are different in many ways. The pro game is longer, involves more shooting and is played at a faster pace than the college game. The games are more likely to turn into track meets in the pros than in the NCAA. Also, the college tournament is a series of “best of ones” played at neutral sites, so it’s easier to just pick a flat-out winner of each game using a formula that doesn’t contain anything involving home-court advantage.

While these items are of some concern, a.) there’s nothing I can really do about it and b.) the formula has effectively projected at least one of the two teams to appear in the NBA Finals in each of the last four years. So, it doesn’t appear to be a complete lost cause.

Anyway, here are the quick and dirty NBA picks, along with my even dirtier (I guess?) guesses of how long each series will take. I sort of have a system when I pick these numbers, but let’s just say it’s in an experimental stage right now.

Read the rest of this entry »

Leave a Comment

Told You So: Kansas and Memphis To Play For National Championship

Exciting times here at Merrill Park Superstar headquarters, as with wins by the Memphis Tigers and Kansas Jayhawks in Saturday’s National Semifinals, the wacky formula that we’ve been working on throughout much of the 2007-08 college basketball season is actually starting to look like a pretty useful tool for predicting the likely outcomes of college games.

The wacky formula, which as press of time lacks a more clever name or even a more descriptive name, has correctly projected the Final Four (not a big stunner) and the teams that are to compete in the National Championship game on Monday. It also had Davidson reaching the Elite Eight. The big issue I found with that bracket, which you can view here, is that the formula called for a lot of first round upsets that didn’t pan out.

I actually kind of saw that coming before the tournament began, though, and decided that I needed to modify the formula a bit. As mentioned in this post, I noticed last year that the Pythagorean theorem of basketball was fairly reliable in predicting the tournament. Whereas the wacky formula figured North Carolina and Kansas would end up in the finals, Pythagoras did a little bit better, forecasting a Florida-Texas A&M showdown. The Gators, of course, went on to win their second straight NCAA Championship a year ago.

Read the rest of this entry »

Leave a Comment

March Madness: The Final Speculative Post

We’re just hours away now from what, in my opinion, is the best two days in sports — the first round of the NCAA tournament. The big question, as always, is who the hell is going to win this monumental playoff? The prevalent choices right now seem to be North Carolina and UCLA — the number one seeds in the East and West regions, respectively.

As I noted in my last post, the wacky formula I drew up and tweaked continually throughout the year in an effort to “crack” college basketball did just okay in terms of picking the winners of the conference tournaments. All in all, we were on the right track — correctly forecasting the winner outright or the sleeper team that won an automatic berth — in 14 of the 30 tournaments and finishing with a “winning percentage” that was a shade under 70. Is this by any means great? I’d have to say no.

Still, when applied to the entire college basketball season to this point, the formula appears to have some relevance. To wit, of the 65 teams in this year’s tournament, it lists the top three teams as Kansas, North Carolina and Memphis. (Trendy sleeper Davidson is listed fourth, but is followed by Tennessee and UCLA.) Of course, it should be noted that the metric doesn’t do anything to take into account schedule, or where games are played. Thus, as demonstrated to some extent by Davidson — some may argue Memphis as well — the obvious flaw with this metric when analyzing a tournament such as this one is that a team can be rewarded for playing in a conference without any legitimate challengers to its dominance.

In an effort to at least somewhat neutralize the effect of a quality team dominating a bad conference, I decided I would integrate the Pythagorean theorem of basketball into the equation. Essentially, what I’ve done is taken the number produced by the old, wacky formula and multiplied it by a team’s Pythagorean winning percentage. The idea behind this is twofold.

First, Pythagoras tends to be a pretty good forecaster in terms of basketball. Last year, it correctly predicted an NBA championship for the San Antonio Spurs and a return to the NCAA championship game for the Florida Gators. That’s not to say it’s perfect — the Gators were to be upended in the title game by Texas A&M, while the Spurs were to defeat the Chicago Bulls in the NBA Finals. Still, the metric has been on the right track more often than not of late.

Secondly, it takes into account margin of victory. If a team had a weak schedule, but won most of its games in blowout fashion, Pythagoras rewards them. Similarly, if, for example, Davidson lost badly in some non-conference games, the formula would reflect that and detract from any blowouts in conference play. Again, it doesn’t adjust for strength of schedule, but it does take into account strength of victory, which the original formula does not place enough emphasis on.

With all of this in mind, I decided to fill out a bracket with the “straight formula” and another bracket with the Pythagoras-modified version. My hope was that it’d help me find some teams with Cinderella potential.

First, we’ll look at the “straight formula” bracket. The bracket, I should note, is stolen from Facebook, because that’s the type of quality operation I’m running here.

formulabracket.jpg

In the first round alone, this bracket contains 12 upset selections (based on seeding). They are:

EAST REGION:
#13 Winthrop over #4 Washington State
#11 St. Joseph’s over #6 Oklahoma
#14 Boise State over #3 Louisville
#10 South Alabama over #7 Butler

MIDWEST REGION:
#9 Kent State over #8 UNLV
#11 Kansas State over #6 USC
#14 CS Fullerton over #3 Wisconsin
#10 Davidson over #7 Gonzaga

SOUTH REGION:
#14 Cornell over #3 Stanford
#10 St. Mary’s over #7 Miami (FL)

WEST REGION:
#12 Western Kentucky over #5 Drake
#11 Baylor over #6 Purdue

Yes, that’s three #14 seeds that are projected by the formula to take their first round games. Historically, the 14-over-3 upset has taken place 15 times in 92 tries — just over 16 percent of the time. In other words, it tends to happen roughly once every two years. So, three such upsets this year seems unlikely.

Furthermore, when a #14 seed does advance, it rarely gets much further. Only twice in 15 tries has a #14 seed followed up a first-round upset with another one in the second round — potentially tough news for this projection of Boise State.

The hope was that weighting the formula with Pythagoras would sift through some of these unlikely upsets (Cal Fullerton over Wisconsin?) and produce a more realistic set of potential Cinderellas. So, did it?

pythbracket1.jpg

Sure enough, two of those three 14-3 upsets disappeared, surprisingly leaving just the Cal-Fullerton upset of Wisconsin. Here’s the entire list:

EAST REGION:
#11 St. Joseph’s over #6 Oklahoma
#10 South Alabama over #7 Butler

MIDWEST REGION:
#11 Kansas State over #6 USC
#10 Davidson over #7 Gonzaga
#14 Cal-Fullerton over #3 Winconsin

SOUTH REGION:
#10 St. Mary’s over #7 Miami (FL)

WEST REGION:
#12 Western Kentucky over #5 Drake
#11 Baylor over #6 Purdue

That looks a little better. And only three of those teams is projected to move deeper into the tournament, with Davidson projected to reach the Elite Eight. Ultimately, though, it’s all four top seeds left standing according to this projection — not bloody likely.

With regard to Davidson, on one hand this formula has actually managed to be pretty reliable in terms of picking upsets and sleeper teams during the year. On the other hand, I’m not sure I completely trust Davidson because of the team’s imbalance. Over the past few years, I’ve made a habit of picking against tournament teams that lean heavily on a single scorer, particularly if the majority of those points come from outside the paint. The thinking is simple: if that one scorer happens to be off, the team is pretty much screwed. This is why I picked against Duke when it was the J.J. Redick show and Syracuse after Gerry McNamara carried the Orange through the Big East tournament. (It also helps that I hate both of these teams).

And yet, despite the fact that Stephen Curry is responsible for nearly 32 percent of Davidson’s scoring, I don’t have much of a problem picking Davidson to reach the Elite Eight. Curry has been stunningly consistent over the past two years — he’s only a sophomore — averaging 20+ points per game to go along with just under five rebounds, 2.8 assists and 2.8 turnovers per game in both of his Wildcat campaigns. If Curry falters — his season-low 14 points came on January 16 against Furman — Davidson will likely have to rely on senior point guard Jason Richards, who averages 12.6 points and eight assists per game, to keep the school’s nation-leading 22-game winning streak alive. (Where have I heard that before?)

There are a lot of interesting questions to be answered in this year’s tournament: How long can Davidson keep its streak alive? Has Georgia turned the corner at just the right time? Can George Mason revisit its success of two years ago? Can Memphis prove its doubters wrong? Will this be another big year for the mid-majors?

And there are plenty more to be asked. Luckily, we won’t have to wait much longer to learn the answers.

We’ll be back routinely throughout the next couple of days to cover some of the bigger stories of the early rounds of this year’s NCAA tournament.

Leave a Comment

If Our Formula Was A Basketball Team, It May Have Been On The Bubble

I intend to join the whole bracket madness fray a little later with regard to my picks, and sleepers, and so on. But I figured before I did that, it’d make sense to take a look at how the formula I used to pick the conference tournaments fared.

Read the rest of this entry »

Leave a Comment

Amateur Bracketologist: How Bad Did We Do?

Final projection! Congratulations to Georgia on inexplicably winning the SEC Tournament.

east2.png
west1.png
midwest1.png
south1.png

FINAL FOUR:
East Regional Winner vs. West Regional Winner
South Regional Winner vs. Midwest Regional Winner

MULTIPLE-BID CONFERENCES:
SEC: 8
Big East: 7
Big 12: 6
Pac 10: 5
ACC: 4
Big Ten: 4
Mountain West: 3
West Coast: 3
A10: 2
Sun Belt: 2

LAST FOUR IN:
Oklahoma
Arizona State
USC
Kentucky

LAST FOUR OUT:
Saint Joseph’s
Massachusetts
Oregon
Virginia Tech

Leave a Comment

Amateur Bracketology Revisited: Slightly More Pleasing To The Eye Edition

As promised, here is my latest series of Photoshop masterpieces. The brackets are updated as of the conclusion of the Texas-Arlington and North Carolina wins, though you won’t actually see any changes since I had projected these outcomes prior to creating last night’s projections.

Just give the brackets a click for a better look. We’ll try to get back prior to the Selection Show if anything unexpected (like, say, Georgia winning) happens.

east1.png

west.png

south.png

midwest.png

FINAL FOUR:
East Regional Winner vs. West Regional Winner
South Regional Winner vs. Midwest Regional Winner

MULTIPLE-BID CONFERENCES:
Big East: 7
SEC: 7
Big 12: 6
Pac 10: 5
ACC: 4
Big Ten: 4
Mountain West: 3
West Coast: 3
Atlantic 10: 3
Sun Belt: 2

LAST FOUR IN:
Arizona State
Southern California
Kentucky
Saint Joseph’s

LAST FOUR OUT:
Massachusetts
Oregon
Virginia Tech
Minnesota

Leave a Comment

Amateur Bracketology Revisited: Quick and Dirty Sleep Deprived Edition

I couldn’t sleep, so I figured that instead of counting sheep, I could take another crack at what the bracket is going to end up looking like before its release on Sunday.

I was looking for a graphical version that I could just fill in the names and upload, but it looks like it might be a little more complicated than that. So, while I might bust out Photoshop tomorrow in a vain attempt to make my bracket look prettier (it may be the no sleep talking, but I’ll probably update my bracket throughout the afternoon if need be), right now it’s going to be all text, all the time.

Read the rest of this entry »

Comments (2)

Older Posts »