First, a congratulations to the Tigers of LSU as they have taken the inaugural SEC Baseball Pythagorean Power Ratings Championship.  They won by the slimmest of margins, taking out Florida by a total of .002 percentage points.  Way to go Tigers, you are champions of math!

As you may remember, I have been rating teams all season long based upon their ratio of runs scored/allowed to give an indication of how many games a team should have won or lost in comparison to their actual record.  For the most part this system, developed by baseball statistician Bill James many years ago, predicted a teams record within one or two games in many cases.  This remained true as eight teams finished within two games of their predicted total. There were a few anomalies like Auburn who lost way more than they should have but for the most part it was very accurate.  Since it is accurate in the regular season I wanted to test the formulas accuracy in SEC Tournament play.  This is dangerous because in tournament play one game can ruin an entire samples worth of data, but I wanted to see if this system could predict SEC Tournament games better than the actual seeding.  Below you will find chart of the Final SEC Pythagorean Standings, Log5 Predictions, and Seeding Predictions for this weekend’s tournament.

The first match-up we’ll examine is the 4 vs. 9 game: Kentucky and Ole Miss.  The seeding implies that this game should be a relatively easy win for our ‘Cats, but Log5 says this is not the case.  As of late Kentucky has been falling in the ratings while the Rebels have been steadily climbing.  I anticipate that this early morning match-up will feature a fair amount of offense as Kentucky and Mississippi both have great offenses and average defenses.  Log5 gives the Wildcats a 54% chance to win and I tend to agree as I see Kentucky advancing, but it should be close.

The second match-up we’ll examine is Georgia vs. Vanderbilt.  The seeding predicts that this game should be competitive but Vanderbilt being several runs better. Log5 disagrees, as it has predicted the tightest game of the day giving the Commodores a 52% chance to come out victorious over the Bulldogs.  I anticipate a balanced offensive and defensive game for both teams with Vanderbilt coming out on top in a close one.

The third game up for examination is Mississippi State vs. Arkansas.  The seeding says that this game should be the closest of the day as #6 takes on #7, but again Log5 disagrees.  Through the course of the season Arkansas has proved themselves the better team and should play like it tomorrow.  Both teams are excellent on defense, giving up fewer than 200 runs on the season, but Arkansas’ offense is significantly better than Mississippi State’s.  I look for Arkansas to win this one by a convincing margin and prove that they are better than their seed indicates.

The final game for examination is Auburn vs. Florida.  The seeding says that this will be a blowout and Log5 agrees for a change.  Florida is the second best team in the league according to the Pythagorean Power Ratings (just decimals away from LSU).  Auburn is the most unlucky team in the tournament, losing a full five games less than they should have, so maybe their luck can change against the Gators.  This is unlikely as Florida is elite offensively and defensively so the potential for upset is slim.  Log5 has given the Gators a 64% chance of winning and I agree.  The Gators win this one going away.

To conclude, while the outcomes are predicted to be the same in the seeding and Log5 the way those outcomes are predicted are totally different.

Follow me on Twitter @SchuetteKSR