The following is an argument I had with a fellow Red Sox fan Ben Gritz over the worth of Mark Bellhorn. The 4 short discussions occurred starting late in the 2004 regular season and are dated to the best of my ability. (Note: Ben would like to point out that he no longer fully endorses all of his work below.)

Mark Bellhorn, the Man Behind the Mask
By John Williams, Concerned Member of Red Sox Nation (September, 2004)

Anecdotal memory seems to suggest that Mark Bellhorn gets an inordinate number of hits per ball put into play (ie any ball that requires fielding). He has an undeniable knack for walking and striking out, but I am hypothesizing that Bellhorn would have a significantly lower batting average this year if not for his luck when he actually does hit the ball in play. I believe that Bellhorn’s reason for being in the lineup is for his ability to get on base, but I think he would get many less starts, and not bat 2^nd in the lineup sporting a low .200s average, which is what kind of hitter I believe he is. Looking at his statistics this year as of 9/21/04, he has had 490 official at bats and 128 hits (including 17 home runs). Subtracting out the at bats in which he has homered and the at-bats where he has struck out, it leaves 111 hits in those other 312 at bats. This is a “hit percentage” of 35.6%. Over his entire career (1507 at-bats, 362 hits, 53 homers, 500 strike-outs), he has 954 homerless and k-less at-bats and 309 hits that aren’t homers. This is a hit percentage of 32.4%. This difference in percentage is manifested in the fact that Bellhorn is batting 0.261 this year, 20 points higher than his career average of 0.241. These career numbers, though, are weighted heavily towards this year because it is only his second full-time season. Looking just at his career before this year, he has 1017 at-bats, 234 hits, 36 homers, 339 k’s. This leads to 642 homerless/kless at-bats, and 198 hits, a percentage of 30.8%.

A good comparison to make is to a hitter like Ichiro, who has the speed to beat out almost any ground ball he hits, leading to what should be a lofty hit percentage over his career. Looking at his numbers over his mlb career, he has 2672 at-bats, 905 hits, 37 homers, 240 k’s. 2395 at-bats and 868 hits. This is an average of 36.2%, quite similar to Bellhorn’s hit percentage this year, so either Bellhorn has suddenly developed ichiro-like speed, or he is indeed having a lucky season. Looking at all of the red sox hitters this year (a crude way to take into account park and schedule effects), they have had 5279 at-bats, 1487 hits, 205 homers, 1098 k’s. 3976 at-bats, 1282 hits, 32.2% hit percentage, very similar to career average.

Now let’s ask “what if.” To me, bellhorn seems slower than pujols but let us just say that pujols and bellhorn are of similar speed and thus should have a similar hit percentage. Pujols this year has a hit percentage of 29.6%. Now, giving Pujols bellhorn-like luck in his 463 k-less and homerless at-bats would lead to 165 hits in those at-bats. Adding up his totals for this year with that hit percentage would lead to 553 at-bats, 209, at batting average of 0.378. Basically he would be Barry Bonds but with less walks. How about Barry himself with Bellhorn’s luck? A batting average of 0.397, ladies and gentleman, Ted Williams has just entered the building.

My main point is that I do not think it is unreasonable that during the playoffs we should expect a hit percentage of about 32% for bellhorn. Figuring that hit-percentage for his year this year, he would be batting 0.238 (using Bellhorn’s career hit percentage not including this year, it would be a batting average of 0.226). I, for one do not want a 0.238 hitter who strikes out more often than most sluggers batting second in the Red Sox lineup in the playoffs, especially since he lacks nearly every other baseball skill (ie bunting, making contact on hit-and-runs, range in the field, reading pitches).

Counter-arguments, criticisms of methods, or pointing out of math errors would be welcome.

Mark Bellhorn, Chicks Dig Him
By Ben Gritz, sole supporter not named Epstein (September, 2004)

The biggest problem with evaluating Mark Bellhorn is the fact he simply has not had many full years of major league service. There’s a bit of a problem with the sample size in any Bellhorn projection. And once you consider that the playoffs are themselves a small sample size, predicting his future or playoff success with any objective certainty is doubly difficult. Here are his lines for the only two seasons in which he’s had significant service-time:

2002 – Chicago Cubs

2004 – Boston Red Sox (projected on ESPN.com)

Some of his numbers are consistent and many are not. AVG and OBP however are eerily similar, which would indicate to me that Bellhorn performs better (or “luckier”) if he has more ABs, although a third full-year would be useful. Now, John charges that Bellhorn is a much lower batting-average hitter than the bare stats indicate. Even if we pretend that his career average of 0.240 is his batting average for this season, the difference is only that of 11.28 hits; which is roughly an extra hit every dozen or so games. And even if we take a “worse-case” scenario where we drop his average to his career number prior to this season (which John cited as 0.226), the number of actual hits lost is still only 18.672, or an extra hit every week. Considering the strength, and more importantly, the consistency of his OBP (career 0.354) I still am willing to trade away his possible “lucky hitting” in favor of getting to first and beyond.

The bottom line is that due to sampling restrictions it’s hard to demonstrate whether Bellhorn is a useful long-term everyday player. He’s probably not. But, let’s look at how he’s done this year as contrasted to two other groups of players: second basemen and #2 hitters. This I think better illustrates his value to the Sox this year (as well as his place in the lineup) than looking at a relatively small sampling of hit percentages.

Here are his ranks amongst MLB second basemen in a number of important offensive categories as of games started on 9/26 (according to qualified 2B on ESPN.com).

tied for 3^rd in Runs, 13^th in Hits, 4^th in Doubles, 5^th in Homers, 3^rd in RBIs, 1^st in Walks, 16^th in AVG, 3^rd in OBP, 4^th in OPS, 4^th in Runs Created, 3^rd in Runs Created/27 Outs, 1^st in Secondary Average (looking at bases gained independent of AVG [formula: (TB - H + BB + SB - CS)/AB], 6th in Extra Base Hits

Bellhorn ranks in the middle of the league or higher in every significant offensive statistic among second basemen. And if we ignore hits and AVG (two categories paid less attention to by good ole Moneyball guys) Bellhorn looks like just about the best 2B in the league whose name doesn’t end with Loretta. I’m particularly impressed by his doubles, homers, walks and OBP/OPS. Obviously strikeouts are a serious problem not shown here; although interestingly he’s still in the middle of the 2B pack when it comes to BB/Ks despite all the punchouts. But between the stats I just listed I would say his offense drowns out his strikeouts and average.

What about defense!? Now, I don’t care much about defense (my stance on the Nomar trade is sufficient evidence for that), but let’s take a gander at El Marko. Bellhorn has 11 errors at second, fewer than illustrious offensive 2B with names like Soriano, Boone, Belliard, Womack, and Hudson. The only significantly good offensive 2B with better error numbers (Kent, Loretta, etc.) are within 4 errors or less of Bellhorn. Of course errors aren’t hugely indicative beyond immediate fielder-competency, and to his discredit Bellhorn’s positional fielding percentage is near the bottom of the league and his double plays/games started ratio is 0.531; which isn’t that great league-wide but comparable to Jeff Kent’s score of 0.538. His zone rating however is 3^rd best in the league. Again, not a great defensive player but a competent fielder who drowns his defense with offense.

Now, for the coup de grâce. Let’s look at how he does contrasted against other #2 hitters on good teams. What I’m going to do is grab every “regular” second hitter (on ESPN’s lineup page of PA leaders at every spot in the order) for every team within 6 games of a playoff spot. If another hitter is within 40 PA I’ll include him too, as well for Beltran. Our nominees are…

Derek Jeter, Mark “He Hate Me” Bellhorn, Christian Guzman / Jacque Jones, Eric Byrnes / Mark McLemore, Chone Figgins / Darin Erstad, Hank Blalock, Marcus Giles, Edgar Renteria, Cesar Izturis, Michael Tucker / JT Snow, Mark Loretta, Corey Patterson, Adam Everett / Carlos Beltran, Placido Polanco

Since we’re looking at production from the hitter position, I’m going to average any numbers from the teams with two guys sharing that hitting spot ([Figgins+Erstad] / 2 for example). OBVIOUSLY, Bellhorn strikes out a ton (in fact only Adam Dunn waves his bat at air more), so that’s a knock against him. But let’s look at some other statistically important areas, and I’ll even still throw in Ks for you. I’m going to grab AB, Runs, Doubles, Homers, RBIs, Ks, GIDPs, AVG, OBP, OPS, RC, RC27 as our categories. Additionally ESPN has included the OPS of the hitter with the most PAs in the 2-hole, so I’ll include that where applicable. Here is my box score:

	AB	Runs	Doubles	Homers	RBI	K	GIDP	AVG	OBP	OPS	RC	RC27	#2 OPS
Jeter	627	107	41	23	76	95	19	0.289	0.350	0.817	101.3	5.63	0.795
Bellhorn	505	90	36	17	80	169	8	0.261	0.372	0.815	88.1	6.18	0.822
(Guzman+Jones)/2	559	82	30	8	45	62	15	0.277	0.311	0.699	64.7	3.98	X
	546	69	22	24	80	115	12	0.256	0.316	0.748	71.5	4.48	X
Minnesota	552.5	75.5	26	16	62.5	88.5	13.5	0.267	0.314	0.724	68.1	4.23	X
(Byrnes+McLemore)/2	541	87	37	20	73	104	11	0.288	0.352	0.831	92.5	6.21	X
	246	29	14	2	21	32	4	0.252	0.360	0.693	32	4.48	X
Oakland	393.5	58	25.5	11	47	68	7.5	0.270	0.356	0.762	62.25	5.345	X
(Figgins+Erstad)/2	550	77	21	5	58	91	6	0.289	0.345	0.753	80.4	5.14	X
	471	75	28	7	65	66	9	0.299	0.346	0.752	67.2	4.3	X
Anaheim	510.5	76	24.5	6	61.5	78.5	7.5	0.294	0.346	0.753	73.8	4.72	X
Blalock	595	102	37	32	109	141	13	0.274	0.355	0.884	105.6	6.27	0.862
Giles	358	59	22	7	43	67	6	0.307	0.377	0.804	61.4	6.18	0.818
Renteria	562	82	36	10	71	73	14	0.290	0.330	0.737	73.9	4.54	0.735
Izturis	646	87	30	4	59	70	6	0.293	0.332	0.718	85.7	4.76	0.721
(Tucker+Snow)/2	445	76	21	13	59	100	5	0.258	0.356	0.776	71	5.51	X
	329	59	30	11	58	57	5	0.328	0.433	0.959	76.4	8.89	X
San Francisco	387	67.5	25.5	12	58.5	78.5	5	0.293	0.395	0.868	73.7	7.2	X
Loretta	600	107	47	16	75	44	10	0.340	0.395	0.900	118.9	7.5	0.893
Polanco	478	69	20	16	48	38	13	0.291	0.338	0.771	68.8	5.04	0.77
Corey Patterson	603	89	32	24	70	160	7	0.272	0.325	0.786	90.8	5.32	0.713
(Everett+Beltran)/2	384	65	15	8	31	56	4	0.273	0.317	0.703	50.3	4.38	X
	313	68	16	23	53	53	4	0.259	0.371	0.939	72.3	8.06	X
Houston	348.5	66.5	15.5	15.5	42	54.5	4	0.266	0.344	0.821	61.3	6.22	X

I'll analyze it for you. Here are Bellhorn's "ranks" out of 14th:
AB - 9th, Runs – 4th, Doubles – tied 4th, HR – 4th, RBI – tied 2nd, K – 14th, GIDP – 4^th(fewest GIDPs), AVG – 14th, OBP – 4th, OPS – 7th, RC – 6th, RC27 – tied 6th, #2 OPS – 3^rd

There are a few interesting things to pick out of here in Bellhorn’s favor. He’s not the only high strikeout guy hitting second as both Corey Patterson and Blalock whiff frequently (Patterson only 9 fewer Ks than Bellhorn). His offensive numbers seem to be at the mid-upper echelon of #2 hitters for all the “good” teams. Considering he’s in the bottom quarter of at-bats for all the players sampled, I think the fact that he’s easily able to keep pace in many of the categories is quite impressive. Here are the numbers Bellhorn would theoretically amass if he had as many chances hitting as say, Jeter, who has an almost identical OPS:

	AB	Runs	Doubles	Homers	RBI	K	GIDP	AVG	OBP	OPS	RC	RC27
Bellhorn	505	90	36	17	80	169	8	0.261	0.372	0.815	88.1	6.18
Jeter's numbers	627	107	41	23	76	95	19	0.289	0.350	0.817	101.3	5.63
"per At Bat modifier"	X	0.17822	0.07129	0.03366	0.15842	0.33465	0.01584	X	X	X	X	X
w/ Jeter ABs	627	111.74	44.70	21.10	99.33	209.83	9.93	X	X	X	X	X

The percentage-related stats (AVG, OBP, etc.) theoretically wouldn’t change, and I’ve spent too much time on this already to bother recalculating Runs Created. Presuming the projection I used isn’t flawed, if Bellhorn had Jeter’s at-bats he would lead or be next best among the league’s #2 hitters in Runs, Doubles, RBIs, and of course, punchouts and walks. It is likely that his OBP/OPS gets a bump as well, and his HR are near the top.

In conclusion, an overall statistical analysis of Bellhorn’s production this year would lead me to conclude that he’s one of the league’s best hitters in the two-hole. Peter Gammons stating that Marky Mark is the “league’s best situational hitter” (on Baseball Tonight), something which I don’t have the time or patience to statistically explore, enhances this conclusion. His adequacy in the field (he’s either really good or highly mediocre depending on the stat selected) is more than made up for by his offensive production, which is good both as a number two hitter and which are great for a second baseman this year. Perhaps best of all is Bellhorn’s salary: $490,000. Is he a long-term solution at second base? Probably not for the Red Sox. Would I prefer to see Bill Mueller at second and the Greek God of Walks hitting in Bellhorn’s spot instead? Of course. But given the current necessity to pack the 5-7 slots with as much offense as possible, Mueller’s balky knees, Pokey’s miniscule bat, and Bellhorn’s incredulous offensive production; it is my opinion that there is no greater positional bargain in the league this season than Mark Bellhorn, and he deserves to start there for the remainder of the season.

Apologia Bellhornica
by Ben Gritz (July 26th, 2005)

This analysis got interrupted by a random trip to the Kells and the Model, but let's see if it remains coherent all the way through...

It's been demanded of me by other baseball fans that once again I must defend my monstrous creation, Mark "whiffs at whiffle ball" Bellhorn. He strikes out almost 40% of the time, and his average is beginning to hover near the Mendoza Line. My god, the horror! The abominable spectacle! The insult to o jogo bonito!

but hey, let's look at some numbers.

this is what Mark Bellhorn did last year, when I (rather soundly I think) concluded that Bellhorn was one of the three best 2B last year, and easily the best for your money:

2004
AB R H 2B HR RBI SB BB K AVG OBP SLG
523 93 138 37 17 82 6 88 177 .264 .373 .444

the on-base children, the on-base!!

anyway, he's had a more rough time this season hitting ninth as opposed to second. This is his line before he went down with that wrist-injury a couple weeks ago:

2005
AB R H 2B HR RBI SB BB K AVG OBP SLG
283 41 61 20 7 28 3 49 109 .216 .328 .360

and did you know that he has an ugly face??

since this line is an incomplete sample, a good way to look at things is to sort of prorate out his numbers. What I do is project what his numbers this year would look like if he had last year's at-bats. AVG OBP and SLG would stay the same in theory because they're percentage stats to begin with. Anyway, this is what FutureHorn would look like with 2004 ABs and 2005 production:

The Not-So Distant Future:
AB R H 2B HR RBI SB BB K AVG OBP SLG
523 76 113 37 13 52 6 91 201 .216 .328 .360

what is most sickening is that his strikeout rate has jumped up alarmingly this season, if you thought he struck out most of the time last year he's much worse this year at this point. But, there are some other things I'd like to point out.

In terms of runs, he's on pace to score a lot more for us than Orlando Cabrera would be projected to had he been on our team all season with Mark's AB total (75.77 vs 56.88), and at over half the salary to boot. Steals are the same, RBIs are down but he is hitting 9th, and slugging is off.

If you look at his power numbers (the ones I really care about), what is amazing is that despite his "struggles" he's on-pace to smack as many doubles as he did last year, and his home run pace is almost the same (which I think considering he hit almost half his homer total in the last month was bound to improve even more anyway). His walk numbers are also slightly above last-year's pace, which is a very encouraging sign considering his struggles. More importantly than that, he's on-pace for 91 walks despite hitting ninth for the team, despite that everyone knows to throw him strikes and when he doesn't have Manny Ortez hitting behind him.

So other than the Ks (and honestly I'd rather he got strikeouts than ground outs) the real problem, the one people probably notice the most since it's flashed more than once every game, is his batting average. I agree, 0.216 looks really awful. But how bad is it? Right now he has 61 hits in 283 ABs. What's a good batting average that people like? 0.275? If you wanted that, Bellhorn would have needed to get 78 hits on the season, which is 17 more than he has now. If you distribute it over the season, not counting the games since he got hurt, it's basically ONE extra hit a week, 1.06/wk to be exact. And since even _I_ don't consider him a .275 hitter, he'd only have needed 9 extra hits on the year to be hitting at a decent 0.250 clip. Considering how many hard liners he was hitting right at fielders earlier in the season (which renowned Bellhorn-hater Johnny J even noticed) I don't think an average of .250 or more is beyond question.

The only thing that makes me even wince at Bellhorn (besides watching him in the box), is that his OBP is only .328, although it is higher than 12 _team_ averages. But considering how low his batting average is, it makes sense, because if he got 9 more luck hits than his OBP would be in the mid-0.360s. And personally I think he had been playing better defense than last year until about the last two weeks, around when his injury was a problem also.

Anyway, while the Red Sox feel that he needs to get a freshen his mental and physical approach to hitting, partly why they put him on the DL, I think a lot of what ails Bellhorn is just what happens to lots of hitters, luck. And luck tends to even itself out over the course of a season, so unless this Graffanino guy we traded for ends up taking over the 2B job, I think hatred of Bellhorn is not quite as merited as the statistics indicate.

Two Points: A. Bellhorn is not a good hitter, and B. Who cares?
by John Williams (July 27th, 2005)

Generally, to anyone who feels that Gritz must defend the fact that Mark Bellhorn is our number 9 hitting second baseman, I would challenge you to suggest some alternatives, because even though he sucks a good amount, it doesn't matter because there aren't better alternatives in our system and he's just the 9 hitter, and he plays his position well enough to not be the biggest problem for the Red Sox.

Now, Ben's email and my boredom has led me to continue the argument that started during last season. Ben says in his recent email "this is what Mark Bellhorn did last year, when I (rather soundly I think) concluded that Bellhorn was one of the three best 2B last year, and easily the best for your money:" Now it's true Ben did accurately conclude this. And why shouldn't he have? .373 OBP, .444 SLG from a second baseman in the AL last year was great, and even better for a guy not making much money. Most people who know anything about baseball could easily look at those numbers and say he was one of the best at his position, there's really nothing to argue about his production last year. A .373 OBP and .444 SLG is good, and good for the future, if you are likely to maintain it. The problem is Mark Bellhorn is not likely to consistently put up those kinds of numbers, taking into account more important trends..

Bellhorn's average last year was .264, which is perfectly acceptable for a #2 hitter given his OBP. A large part of OBP is AVG though, and a player does not completely control his ability to bat for AVG. The argument last year that I proposed was not my own, but one that has been extensively used by baseball analysts: the fact that of balls put in play (not counting home runs), most hitters will see around the same percent fall in as base hits. The batter's ability to affect this percentage mainly is a product of their speed, or their ability to beat out infield hits. There are other factors, but the main point is that hitters are most likely to maintain a hit percentage close to their career average, which is also close to the average of all MLB position players. The statistical product that baseball players are most able to control is in fact their AVG subtracted from their OBP. This number will stay very constant for baseball players from year to year throughout individual careers. Bellhorn, due to his undeniable ability to walk a lot, posts OBP-AVG of .100 to .110 most years, a very good number for most players. I was most concerned last year with the fact that Bellhorn was going to bat 2nd for us in the playoffs, and was most likely going to have a hit percentage close to his career number, and thus bat close to .225 or whatever I concluded, and thus have an OBP in the low .300s, which I did not think would help us in the postseason. I thought it made much more sense to bat a player more likely to have a higher OBP batting 2nd. For anyone who watched the playoffs last year, they saw Bellhorn bat .191 with a .397 OBP, which is an OBP-AVG that is completely unrealistic to maintain over any significant period of time. His postseason performance showed more the randomness of small statistical samples in baseball than his ability. I agree that he was productive over the postseason, but I would have rather had a player who was more likely to be productive in there taking up the 2nd most at-bats for us. Or I could pointlessly argue that if you take away his home run off the foul pole and home run in Yankee stadium that cleared Matsui's glove by a few inches then he would have not been productive. The point isn't that he got a few luck hits or didn't, it's that you need to look at the likely mean production of players and put the best in the most important spots. That was the gist of me last year concluding (rather soundly I think) that Bellhorn was not likely to continue his rate of production.

To do the math again for last year, Bellhorn's hit percentage was about 35%, compared to his previous career rate or 31%, which is a little above the average rate for MLB, but pretty much what you would expect from a player playing in the kinds of parks he has over his career (big outfields in Oakland and Colorado, and good AVG parks in Denver and Boston). Now in 2005, Bellhorn is batting .216, with a .328 OBP, with a 31% hit percentage. This is precisely the kind of year I would expect for him on average year after year. His OBP-AVG is still right in line with his career numbers, as this is a very consistent stat usually, but his AVG is so low because he strikes out much more than a normal player would, so he is not a very productive player, or one you would want to bat 2nd in an order. His numbers are better than some players, like the horrible offensive player Cabrera, or Christian Guzman, but they're not very good. Gritz proposed that if you add 9 extra hits on the year to his numbers then they don't look as bad. Sure, of course they look better. Add hits to any player and their average rises dramatically.
But it's really not that dramatic, there's not much difference between a .275 hitter and a .300 hitter in one particular year in terms of numbers of hits, but over a career it is significant, it's actually 1 hit per two weeks, so really Bellhorn is pretty significantly down in terms of average. A .275 hitter over a career is most likely to bat near .275 in any particular year, and a guy who will end up with a career average near .230 will likely bat near .230 in any particular year. Some years a few lucky hits will get it up near .260, a few years the balls won't fall in and it will be near .200. The "luck" Gritz talked about evening out over a season is already even for Mark, so unless he's just got a pocket full of lucky charms I wouldn't bet on his average climbing.

My second point is, so what that Bellhorn sucks compared to an all-star player, or a player really worth of what he's making. His .330 OBP doesn't do anything to hurt the Red Sox, provided that he continues to bat in the 9th hole, and he should continue to slug well enough to be a regular 2nd baseman in MLB. Alex Cora is a brutally bad hitter, and Graffanino is not better than Bellhorn, so there's no one out there that is suddenly going to make it apparent that Bellhorn should be replaced at 2nd. (Also it's not really apparent why Bellhorn should replace Graffanino when he comes off the DL). So I don't really care anymore that Bellhorn isn't that good. I would like them to pick up a better option in the offseason, as I wanted last offseason, but as long as we are solid at other positions it's not a big deal. What is a big deal is the batting of Renteria and Millar, and the horrible pitching of the bullpen. Oh yeah, also the bad managing of the pitching staff by Francona.