March 22
The Madness Came and Stayed
Once the NFL wrapped up and I threw out all the tissue containers from when Larry Fitz broke my heart, it was time for Men's College Basketball and preparing for March Madness. This year was my year--there's nothing better than carte blanche.
The first thing I decided to do was to get to know college basketball a little better. What are some key statistics that are more valuable than the rest in predictive analysis? We have all heard analysts and talk show hosts regurgitate the words: shooting, defense, rebounding, and limiting turnovers. But which one is more important than the other? Are there other important aspects not being mentioned that may be equally, if not more important? I created a regression model to get the answer.
By downloading advanced offensive and defensive statistics from sports-reference.com/cbb into Excel, I was able to generate a multiple regression model based upon Margin of Victory as my dependent variable, and a plethora of offensive and defensive statistics as my independent variables. The p-value will indicate if the independent variable has an impact on the dependent variable (MoV). If there is an extremely low p-value, that means we are extremely confident it does have an impact and should be kept in the model. Through removing different variables with high p-values, I was able to get a Adjusted R Squared value that was 0.998297. This means that only .17% of the variance in Margin of Victory can be explained by outside factors or randomness, while over 99% of the variance is explained by the factors I have selected to keep in the model.
Reviewing for Multicollinearity
December 27
Boston Red Sox Latest Move: Craig Kimbrel
The last three World Series victories for the Red Sox have come with a stellar closer, and it is no coincidence (Foulke[19th in WAR for relievers in MLB] Papelbon[7th] Uehara[1st]). Over and over the Royals are mentioned for how they succeed: their bullpen. Having a guy in the back end of the bullpen be able to come in and close games out consistently is under appreciated due to the confidence it gives to the team, instead of having the added pressure of needing to score more runs so the bullpen doesn't blow it. These games that are lost in the 9th are back breakers, and they do not allow a team to carry momentum into its next game. This past year really showed how badly a solid reliever was needed as Uehara logged his lowest innings in a Red Sox uniform in a season at 40 and a third innings, partly due to his wrist injury and how the Sox weren't ahead in the 9th as much as previously. The rest of the bullpen was hard to watch as any lead wasn't safe from the bombs that Ogando was giving up. Kimbrel will not fix the Red Sox as there are more holes than just the bullpen, namely starting pitching of course, but will he step into a role that will give the Sox optimism and confidence of the future? Definitely.
Kimbrel is coming off his worst season in terms of BAA and WHIP, while also giving up his highest number of home runs at 6. But, there is not much reason to panic. Power numbers tend to average out over time as the sample size rises, so while he did not give up 12 long balls like Ogando, this is not a particular statistic to be overly concerned about. Especially for closers, as they do not log enough innings to have a large enough sample size to really gauge if the problem lies with the pitcher or a little bad luck, it is usually a little bit of both. Let's look into some trends that give a better indication of future performance.
The statistic most predictable and consistent year to year are a pitchers strikeouts per 9 innings pitched (K/9). The past three seasons for Kimbrel are as follows: 13.16, 13.86, 13.20. All three, and Kimbrel's whole career thus far has featured high strikeout numbers and it is why he has been so dominant, eliminating the chance to even get a hit by fanning the batter. He is 5th in K/9 over the past three seasons and ranks 4th in WAR for relievers over that time span. To show how important striking out batters is in its relation to success, take the top 15 relievers in terms of WAR over the past three seasons and the next 15 after that. The top 15 contained only one pitcher with a K/9 below 10 (Mark Melancon: 8.36), while the next 15 contained 12 pitchers with a K/9 below 10. This demonstrates having a strikeout closer should be valued highly. More to come...
What We Can Learn:
Today I tried analyzing combine drill performances over the past 10 years of top performers at each drill. I deemed a high end NFL Starter for multiple seasons a success, and counted each for each drill at the running back and wide receiver position. This is not credible in solely determining if a player should be a success in the NFL if he does well in those drills that other successful players have done well in, but it is still interesting to investigate.
Key - 40 = 40 yard time; BP = Bench Press; VJ = Vertical Jump; BJ: Broad Jump; 3C: 3 Cone Drill; 20S: 20 Yard Shuttle; 60S: 60 Yard Shuttle
(#*) - Amount of top performers that still have an prominent opportunity, young backup/first-second year starter
Results:
Running Backs - 40: 7(2*), BP: 3(2*); VJ: 3(5*); BJ: 3(2*); 3C: 5(3*); 20S: 2(6*); 60S: 3(3*)
Wide Receivers - 40: 2(4*); BP: 4(3*); VJ: 1(3*); BJ: 5(4*); 3C: 3(1*); 20S: 1(4*); 60S: 3(1*)
This may not be a science, or even close to it for that matter, but determining which successes in drills are better indicators of success in the NFL could just add another dimension in the selection process and narrow down the key factors of a potentially successful skill player.
With the lack of game tape available to me (4 minute highlight reels aren't the best), I am trying to find trends, correlations, any information that can help me in determining a value of an NFL player. This is just one interest/idea that I felt curious enough to explore and here are my interesting findings:
Important RB Drills: 40 Time (Speed/Explosiveness), 3 Cone (Cutting Ability, Acceleration), Vertical Jump (Lower Body Explosiveness/Power)
20 Yard Shuttle is interesting to monitor with 6 prospects as top performers, we'll see if this has any indication of what is to come. It showcases cutting ability and agility, obviously two very important aspects of the game.
Important WR Drills: Broad Jump (Lower Body Explosiveness), Bench Press (Strength, Endurance)
The 40-time intrigues me the most next on the list because if a receiver cannot explode out of his route and continue his burst, he won't be able to break away from corners or safeties.We don't have to look at the numbers to determine it's importance. Speed is a talent, not something to be taught.
January 6
First things first, I'm a rookie blogger... time to start the hazing.
But before that, lets discuss the big game... a much more pressing matter.
Oregon(-7) vs Ohio State
Total Points: 75
The Ducks are laying a touchdown or 6.5 points on most sites to the Buckeyes as they should through their dismantling of teams ever since their early season loss to Arizona. Since their season low, the Ducks have managed to rattle off 9 straight wins beating teams by an average of 27.44 pts. Not only did 6 of these 9 games come on the road or at a neutral site, their most glaring win was their past beat down of Florida State by 39 points in the college playoff semifinals. I agree that these teams are closer than that, but honestly not much closer. The top ranked Ducks offense seems unstoppable sitting at the second best in the FBS in ppg. Being in a high scoring conference like the Pac 12, you'd think their defense would be inflated as well but not at all. The 29th ranked defense in the FBS only gives up 23 ppg. This 23.1 margin of victory average in a highly competitive schedule playing 7 games against teams in the top 25 AP Poll Rankings, winning 6 of those games and their only loss came to Arizona who they beat down in their second meeting by 38. This is a dominant football team thanks to Chip Kelly at the helm... wait a minute its Mark Helfrich following the large footsteps Kelly left, filling them in nicely.
Oregon Key Statistics:
Away Team (as Fav):
Record: 4-0
ATS: 3-1 (3-1)
O/U: 1-3 (1-3)
TeamRankings Rating: 27.9 (#2)
Non-Conference Games:
Record: 4-0
ATS: 2-2
O/U: 3-1
TeamRankings Rating: 31.1 (#2)
And here comes the further bashing that the SEC continues to take, no thanks to their savior Alabama. The SEC lost once again to an Ohio State team who seem like butter because they've been on a roll (Had to throw a Stuart Scott quote in there). The Buckeyes have managed to rattle off 12 straight wins since their 14 pt loss to VA Tech, which seems like a century ago to this team. It is not the same team clearly, led by third string sophomore QB Cardale Jones. From everything I've seen, he is not playing like a third string. He makes smart reads, and if the play isn't there, he uses his legs and size to run for yards, even running through defenders for more yards. Oregon handled one of the nations better running quarterbacks in Brett Hundley, despite giving up 4 yard/rush, they limited his big play abilities as they scampered to a 42-10 lead over UCLA in the early portion of the 4th quarter. So Oregon is not nervous about Jones' running abilities... a much less advanced scrambler. He has still been smart with the football, impressing the nation in the Sugar Bowl with his intelligence only turning the ball over once on an elite Alabama defense. He'll have another tough test against Oregon, who has an average of 2.2 takeaway/game which is more than Alabama's 1.5. But there's something to say about Ohio State's defense who average 2.3 takeaways/game and is only allowing teams to score 22.1 ppg and 333.4 yards of offense per game. They did just allow over 400 yards of offense against Alabama, a lesser impressive offense than the explosive Ducks. If Ohio State is going to win, they cannot, I repeat cannot turn the ball over. The world saw what Oregon just did to Florida State who turned the ball over 5 times. Oregon does not turn the ball over often so they cannot rely on turnovers to get them through this game. Self-infliction will only dig Ohio State a hole they cannot recover from so overall, if Ohio State does not turn the ball over they have a chance at knocking off the Ducks.
Ohio State Key Statistics:
Away Team (as Underdog):
Record: 4-0
ATS: 2-2 (1-0)
Over/Under: 3-1 (2-1)
TeamRankings Rating: 24.1 (#6)
Non-Conference Games:
Record: 4-1
ATS: 4-1
Over/Under: 4-1
TeamRankings Rating: 20.4 (#7)
The Pick: These teams are gunners and can put up points in a hurry. The speed of Oregon's offense and the tempo they play with show how quickly they can put up points on the board. Ohio State won't shy away, and Jones has shown his big play abilities with his big arm and frame.
Oregon has played a tougher schedule, played better away from home, and with Mariota running the show versus a young Cardale Jones without the same elite quickness, accuracy, or decision making as his counterpart, I don't see Ohio State coming out on top or covering the 7-point spread. Oregon will face its toughest test, but the best team in college football will come out on top 45-35. Oregon +7, Over 75
