Baseball 2010 Discussion

The most important aspect for me in handicapping baseball is finding value, most around here do the same, but it tends to get tricky when we start discussing “how we find value”.  One of the stats I like is BABIP, or Batting Average Balls Put in Play.  This stat can be used with hitters and pitchers, and it’s important to note that the MLB average usually hovers right around .300.  This is one stat that we can use to measure a player’s luck.  ERA can be very misleading, especially when using a small sample size.  How many times have you seen some pitcher with a 0-2 record and era north of 8.00 in his last 3 starts win easily as a +145 dog in that 4^th start?

For me, I like BABIP because it can allow one to find value in either backing one team, or in fading another team.  Most players will regress to the mean over time.  There are some exceptions, for instance, Eric Byrnes has historically been a pop fly hitter, and those pop ups rarely fall in for hits, so his BABIP is typically below major league average.

Looking at last season, there were 139 pitchers that had at least 10 starts before the All-Star Break, and one finds that Dan Haren posted the best BABIP in the majors when he clocked in at .234.  This led to an overall opp. avg of  .189 over that stretch, as Arizona win 11 of  18 games he started.  For me, this would have been fade material in the second half.  Arizona still won 8 of his 15 second half starts, but his opp. avg went up to .267 and his era was closing in on 5.00. 

Scott Feldman offers another example,  he closes the first half with a BABIP of  .241, which led to the Rangers winning 11 of his first 15 starts.  His era was 3.21 and his opp avg was .222 as a starter in the first half.  In the second half, add 40 points to that opp average, more than a point to his era, as the Rangers only won 9 of his 16 starts.

Both Haren and Feldman ranked in the lower half of the leader board when you look at the 122 pitchers who had at least 10 starts in the second half of the season, and both were north of .300 with their BABIP.

If you blindly faded Haren and Feldman in all of their 31 post-break starts, you would have lost 1.23 units, but I think it’s reasonable to think a handicapper would have found 3 to 5 games in which you avoided the auto-fade because of the price and/or other factors.  It could have the Feldman game in Tampa when the Rays were laying 66 cents juice, or maybe it would have been the game in KC when the Royals were laying 32 cents.    Ditto for Haren, would you have laid -154 or -142 to fade him?  Eliminate those 4 games, and that 1.23 unit loss becomes a 3.46 unit profit. 

BABIP can be a decent leading indicator when looking for over valued favorites and under valued dogs.  Certainly, other factors come into play, most notably the price I’m paying.

Throw in your two cents, which indicators help you find value ?

good points, i pay way more attention to righty/lefty splits than i do to batter vs pitcher matchup stats.  imo, those stats are somewhat worthless for predicting future performance, but can be helpful finding good value in a pitcher who has been hit hard by the opponent, or in finding good fade value in a starter laying a ton of juice in part because the matchup stats favor him.

i think this stat has a tendency to inflate the line one way or another at times, thereby creating value for us.  i agree entirely with your philosophy on that stat, but i like the fact it's in every paper in the country.

Pitchers tend to run hot and cold, and there only a few "workhorse" types that can consistently run out there  7, 8, 9 innings with high pitch counts.  There are certain over-rated pitchers who tend to sputter for a start or two after a stellar game with a high pitch count.  Look at Buerhle, after his perfect game, he did not record a win for nearly 5 or 6 weeks!

Quote Originally Posted by jakelv:

Great reading i i agree 100%. I noticed you only played seattle games last year will you look at the whole card for each day or just stick to seattle? 

it's mainly seattle except for the random auto-plays, fade wakefield on the west coast, back rogers in oakland, stuff like that, otherwise it's just the Mariners.  my style demands a fair amount of time, and i right now i'm not organized enough to branch out to other games.

Another important stat for me is LOB %, percentage of base runners stranded.  The MLB average has been just north of 70% for the past several seasons.  When looking at a quality MLB pitcher, one can project regression from a rate that is substantially above that level, thereby giving me value in fading a certain pitcher, or I can project progression from a rate that is very low, thereby giving me value in backing a pitcher with a high ERA.

Justin Verlander offers a great example of this from last season.  In April, Verlander posted the worst LOB rate of the 100 qualifying pitchers in MLB. He stranded only 50.3% of his runners.   It was very unlikely that Verlander would continue to strand runners at that rate.  In 2006, he was above the MLB average for the season when he closed in on 80%, while in 2007 he was at 75%, as expected he regressed in 2008, but only to 65%.  Based on this 3 year sample size, is it reasonable to think he maintains a staggering low LOB rate of 50% ?

Looking back at Verlander in 2009, he could have been backed for pretty cheap prices following his horrible April starts.  And his horrible April can be explained quite easily . . . 50% of his base runners scored, opposing batters recorded hits 40.8% of the time they made contact, his FIP was a good 3.13 (Fielding Independent Pitching, more on this in a later post) while his ERA soared to 6.75. 

Now let’s look at the month of May.  Verlander went 5-0 in 6 starts with the Tiger losing his only ND.  In those 6 starts his average price was -115.  Quite a deal for a pitcher that would finish with 19 wins and 3^rd place finish in the Cy Young balloting.  How did Verlander turn his season around ?  Opposing batters still put balls in play, but now those batted balls found Tigers defenders more often, his BABIP was a very reasonable .292 in May.  And when hitters did reach base, Verlander stranded them 84.7% of the time.  Sure he posted a better K-BB ratio of 5.09 in May, but a 3.78 ratio in April is hardly poor. 

In April, Verlander suffered from bad luck which allowed opposing teams to get base hits at a rate that was unlikely to be sustained over the course of a season, which in turn led to a lower than normal LOB rate.  In my opinion, this created a situation where the average handicapper and/or odds maker undervalued Verlander in May.  And that’s what we are looking for.  A Cy Young candidate laying on average -115 for an entire month while facing such powerhouses as Cleveland twice, Minnesota, Texas, Kansas City, and Baltimore.

I don't mind looking at last three game stats.....you just have to dig in a little and make sure there isn't one game skewing the stats, also see what teams they have faced. Make sure it makes sense. I like looking at lefty/righty stats.....have to pay attention to the bullpens, I think there are only about 11 competent relievers in MLB right now .....Good to know which teams play good/bad defense as well. Great pitchers can be victimized by a team with no offense and/or poor fielding. Lots to look into and I think that's why it's fun, at least it is for me.

I enjoy your posts 3825. Not sure how I missed this one for two days.

exactly . . . that's why i like to take advantage of a price that I think is overly influenced by a small sample size.  there is no way that verlander would allow hits at the rate he was allowing them, or that haren or feldman would limit hits at the rate they were doing so.  eventually the prices tend to catch up with this "new trend", but we can take advantage in the interim.

be easy . . . here's my take on this.  First, I like Pitcher at Ballpark stats in cases in which I either have a large sample size, or a plausible reason as to why a pitcher should be good or bad at a specific ballpark.

for instance, Jarrod Washburn has shown a good ability to pitch on Domes, i think he historically succeeded in that area because he historically has a GB-FB ratio that shows he gets more flyball outs than a normal pitcher.  Assuming he has decent outfielders who can handle the roofs, the turf has no impact on his flyballs, while a GB pitcher is potentially prone to seeing eye singles, balls that would be outs on grass turn into hits in Minnesota or Tampa.

Tim Wakefield is the exact opposite.  He has excelled in domes, but has struggled mightily on the west coast over a pretty large sample size this decade.  In wakefield, i can make a case based on the large sample size, and I can give a plausible reason . . . the air inside the domes is predictable, while the air in every west coast ballpark tends to get very heavy at night, even in southern california.  Anaheim isn't as bad as San Diego or LA, but you can feel it there after the sun goes down.  Wakefield has consistently struggled in all these ballparks, which are all within 10 miles of the Pacific Ocean, and none of these places tend to have that warm humid air that you find at the ballparks near the Atlantic.

another angle would be Carlos Silva, a righthanded pitcher who has historically struggled to get groundball outs vs lefty hitters pitching half his games in a ballpark that only favors left handed flyball hitters.  Safeco favors left handed pitchers, while fenway favors right handed pitchers.  in these cases, I might read more into a small sample size than i normally would, if a specific pitcher conformed to these theories even if he had limited starting or innings pitched data, (between 4-8 starts).

fangraphs is a favored site of mine for saber stats, there claim to fame is their predictor graphs, but last i checked those predictor graphs DO NOT take into account the players playing, they are situational only.  thus every team starts the game with a 50% chance of winning.

a quick glance at Aaron Harang's numbers, and I can't find anything too out of whack.  his BABIP was at a career high last season of .339, but that is not horribly higher than his career average of .319, or the .325 he posted in 2006 when he won 16 games.  it is substantially higher than the .298 he had in 2007 when he went 16-6 though. that explains some of it, also, his K ratio has declined the past two seasons, that explains a bit more. in addition, his HR's allowed per Flyball has also slightly increased over the past two seasons, and he has allowed more flyballs over that same time period.  couple these factors with no improvement in his LOB %, and you get back to back sub par season. in harang's situation, I think a case can be made that age is catching up with him, he turns 32 in May.  in simple terms, he might not be as good as he once was.

re: home-road splits,  again, i need a large sample size in order to make this case for a player changing teams.  with small sample sizes, it could be ballpark factors, or simply bad luck in the key areas of BABIP or LOB%.  ervin santana was a classic example, for a long time, he could not pitch worth a shit on the road. in most cases, i think it's critical that you first consider ballpark/opponent/luck factors before determining whether or not a player is a home or road player.

in fangraphs, you can search for players, as well as leaderboards.  i also use hardball times, but the stats are a bit trickier to deal with.

looking at some of the props on Pinny, and one is in particular stands out . . . highest hit total by a mlb player . . . over-under 218.5.

this one popped of the screen due to my proximity to the leader in hits for the 2000 decade, ichiro.  who posted 225 hits last season in 146 games.  ichiro has been under that total in 4 of his 9 seasons, and in some of those cases, there is a plausible reason.  first in 2002, he was due some regression because mlb pitchers became more familiar with him,  in 2003, i have no excuse, but i think it's fair to say he was on horrible teams in 2003 and 2008, in in the latter season, he was the target of a fair amount of clubhouse adversity.

200 hits is a big deal for ichiro, he set a new record for consecutive 200 hit seasons last year, and he can tie pete rose for most overall 200 hit seasons if he gets there again this season.

i think ichiro has really benefitted from the wak system, he has given some, and wak is smartly letting ichiro play his game, and then trying to use that game to the team's advantage.  ichiro has only been hurt once in his career, and that was the first two weeks of last season.  certainly the 36 year old will eventually slow down, but i think it's fair to project him with better than 218 hits this season based on his 35 year old season of 2009.  no one disputes his preparation, and i think it's plausible that he will lose a step at some point, but he didn't do it last season.  57 of hits were of the infield variety,  5 more than he had in 2008.

looking at team totals,  the total for highest loss total for a mlb this season is 101.5

9 teams in the decade surpassed that dubious mark, but only a single team has done it in the past 4 seasons . . . the nationals in 2008 and 2009.  are we seeing a smidge more parity, or have the past 4 seasons been a slight abberation ?  2002 featured 3 horrible teams, while 2004 had two, and there was one each in 2003 and 2005.

the total for highest win total is 100.5 . . . . ten teams have met this par since 2000, but only one has done it in the past 5 years, the 2009 yankees.

in my opinion, we have correlated trends here, fewer 102 loss teams, and fewer 101 win teams in the latter half of the past decade.