July 13, 2005

While many websites, newspapers, and magazines are currently offering their first half awards and second half predictions, I'd like to take a pass on that because I don't think you need me telling you that Derrek Lee had a great first half and it is somewhat pointless for me to change my preseason predictions just because of the first half results. Obviously, if I could go back in time, I would pick the White Sox to win the division. I won't change my choices, though, because if the Twins win 20 in a row in September and steal the division, then I look like a fool for changing my pick. Predictions are just that, and updating them every time an unexpected event takes place evaporates their usefulness, like a stock analyst downgrading a security AFTER it tanks. What's the point?

Instead, I'd like to drum up the age old debate about a pitcher being worthy of an MVP Award. Regardless of the fact that the Cy Young Award exists solely for pitchers, the question raised here is whether pitchers have enough impact on the game over the course of a season to be considered valuable enough to make up for the games in which they don't play. The first step in doing this is finding a plausible measurement. Specifically, how much impact on a season does a pitcher have compared to an everyday player?

Trying to keep it simple and not involve any "Bill Jamesian" logarithms, we can look at the average number of plate appearances affected by the hitter and pitcher over the course of a season. We will use some liberal rounding but the conclusion should not be affected. Using 2004 season statistics, each game can be broken down into roughly 76 plate appearances. Even an MVP-type hitter is likely to miss a few games during the season (only a couple players each year play in 162 games), so we'll say that the MVP hitter plays in 156 games. At the plate, hitters average roughly 4.2 PA/game, and we'll round up to 5/game because of the minimal effect they may have on other at bats while on base. Obviously, speedy Scott Podsednik has a much greater effect on his teammates' at bats than Frank Thomas, but Thomas gets on base more, so an extra .8 PA/game seems about right.

On the fielding side, the hitter is involved anywhere from 400 to 600 to 1,200 times in a season depending on the position. A first baseman obviously gets many more chances than an outfielder, but his chances are not as critical because many times it involves catching an easy toss as opposed to running down a flyball in the gap. If we use 625 chances per season, we can assume that the hitter has an effect on 4 additional PA/game while in the field. That may sell the first baseman a little short and boost up the outfielder a little more than necessary, but it's as good a number as any while trying to combine all of the positions. A catcher, of course, is much more valuable on defense because of his effect on virtually every pitch, so he should be considered separately. In other words, if Ivan Rodriguez were able to post similar numbers to Manny Ramirez and it could be argued that I-Rod's defense has a positive effect (which it does), then I-Rod is MUCH more valuable than Ramirez because he has a much larger effect on the game while posting the same offensive numbers.

Getting back to the numbers, we have concluded that an MVP-caliber hitter plays in 156 games and has an effect on approximately 9 plate appearances (5 hitting and 4 fielding) per game. Out of 76 PA/game and taking into account the 6 games in which the hitter does not play, the hitter has an effect on 11.4% of all plate appearances during the season (formula: [(9x156)/(76x162)] = 11.4%). Let's look at how pitchers fare in comparison.

In today's era, a starting pitcher with MVP "stuff" (Clemens, Pedro, etc.) averages 34 starts, and that includes working on the occasional 3 day rest. These days, an MVP-type year means a pitcher has averaged about 7.5 innings per start, which means the pitcher has an effect on 83% (7.5/9) of the plate appearances, or about 31.5 of 38 while in the field. National League pitchers also hit about 3 times per game, and American League pitchers hit only in interleague games on the road. Averaging them out and giving some leeway because of the "double effect" a pitcher may have in the field occasionally (once as a pitcher, then again as a fielder on the same play), it is safe to say that an MVP year for a starting pitcher consists of approximately 34 PA/game in which the pitcher affects the outcome. Out of 76 PA/game and taking into account the 128 games in which the pitcher does not play, the pitcher has an effect on 9.4% of all plate appearances during the season (formula: [(34x34)/(76x162)] = 9.4%).

Without rehashing the numbers again, a reliever who has a season like Eric Gagne has had in recent years accounts for only 2.9% of all plate appearances during the season. What do these numbers tell us? For one, it tells us that a closer should NEVER win the MVP Award. A pitcher who throws 80 innings out of almost 3,000 half innings played by a team should never be considered as valuable as an everyday player no matter how many saves are recorded. Every time a closer enters the 9th inning with a 3-run lead, that just means there were a group of guys doing all the right things for the first 8 innings. Imagine a pitcher who had Gagne's '03 stats but only pitched the 1st inning of games rather than the 9th inning: his stats would be the same except for the saves, which is a manufactured stat anyway. For those of you who argue that the 9th inning is so much harder in which to succeed, well, I simply disagree. Mentally, it may be a more challenging inning and is arguably worth a little more weight than what the numbers read, but not to the point where hardly being on the field of play can be overlooked simply because the plays take place at the end.

Now that we have discounted closers, let's look at the relationship between starters and hitters. Hitters affect over 20% more of the plays in a season than starting pitchers (11.4% vs. 9.4%) which is a large enough gap to suggest that the pitcher has to do something absolutely remarkable to even be in contention for an MVP. A simple 20-win season with a 2.80 ERA won't make the cut in my book if there's a hitter popping out 50 HRs, 120 RBIs, and plays above average defense. It can be said that of the 9.4% pitchers are involved, the MEASURE of their effect is greater than the hitters' 11.4%. I don't really buy into that because no matter how good or bad a pitcher is throwing, a large portion of the outcome is tied into the hitter, with another, yet smaller portion, tied into the fielder's ability. Even if we discount the effect of the fielder and decrease a hitter's effect to 8 PA/game (from 9 PA/game), the hitter still carries a 10.1% effect, slightly more than a pitcher. I hesitate to do that, however, because an A.L. pitcher doesn't hit (much), so his effect is slightly less than my calculations, and an N.L. pitcher's effect on the offensive side is likely a NEGATIVE one (even Dontrelle Willis or Mike Hampton will finish with lesser stats than a position player hitting ninth), offsetting any argument that a pitcher has more control with the plays in which he is involved.

My conclusion: for the most part, the writers have properly avoided pitchers as MVPs, especially in the N.L. All of the MVP Awards won by pitchers since 1968 have been in the American League and they have all been mistakes except one. Eckersley won in 1992 despite throwing only 80 innings and, therefore, did not deserve the award based on my numbers. Roger Clemens and Willie Hernandez probably didn't deserve the MVPs in 1986 and 1984, either, but at least Clemens logged over 250 innings and Hernandez pitched almost twice as many innings as today's closers. I don't think that justifies HIS award, but it certainly helps discredit an award won by a pitcher who throws a mere 70 or 80 innings. Rollie Fingers won the MVP in 1981, giving up just 9 earned runs on the season, but he only pitched 78 innings! Vida Blue was the last DESERVING MVP pitcher, logging 312 innings and posting an ERA of 1.82 in 1971, including 24 wins, 24 complete games and 8 shutouts. Blue's achievements truly were valuable enough to warrant an MVP Award. His stats equate to an effect on 11.7% of the plays that year, even more than today's hitters. The same holds true for the MVPs in 1968, Bob Gibson and Denny McLain. Gibson finished all but 6 of his starts (and posted a ridiculous 1.12 ERA in over 300 innings), while McLain made 41 starts and won 31 games. In this day and age, it would take a pitcher like Roger Clemens to continue his current pace and finish with an ERA below 1.50 and over 250 innings to be valuable enough for an MVP Award. Until then, the Cy Young awaits...