m.e.t.b.o.t.

Niese 5.73 Tejada 3.62 Wright 0.39 Rice 0.15 Parnell 0.11

Duda 2.86 Baxter 1.99 Valdespin 1.99 Buck 1.95 Wright 1.06 Hawkins 0.08 Rice 0.06

Familia 0.60 Davis 0.36 Turner 0.24 Edgin 0.12

Tejada 2.98 Murphy 2.94 Hefner 2.10 Lyon 1.51 Cowgill 0.46

batmagadanleadoff wrote: m.e.t.b.o.t. wrote: in this case, metropolitan centerfielder jordanny valdespin hit a walk away big bang home run which resulted in a metropolitan win in the bottom of the tenth inning. superficially, this would appear to have generated a tremendous amount of WPA. however, the bases were loaded with one out, and the metropolitans faced only a one-run defecit.... Is it about contributing to a team win, or about individual accomplishments irrespective of team outcome? And how do you even begin to separate the two? Why did Santana practically* max out the scoring for his no-hitter? By your logic, because the Mets won by eight, Johan could've given up seven runs on 12 hits without jeopardizing the team win. Why then, was he credited fully for a brilliant, but according to your rationale here, superfluous performance? And then why does any Met, for that matter, ever get any Schaefer points in a team loss? * But see my ballot. m.e.t.b.o.t. does not fully understand the question. m.e.t.b.o.t. has one primary purpose - to process per-game win probability values into schaefer votes. a need was determined, and m.e.t.b.o.t. was shoddily cobbled together from readily available scrap material. it was a very low budget affair. win probability values place the players' individual accomplishments into the overall context of a team win. not all home runs are equal. some are completely meaningless in the context of a single game. others can swing a sure loss into a total victory. similarly, not all shutouts are equal. in many cases, the contributions of the offense make less meaningful the efforts of the pitcher. although it could be argued that in throwing a shutout, the pitcher has made less meaningful the contributions of the offense. if it is necessary to revisit the no-hitter thrown by metropolitan pitcher johan santana, then m.e.t.b.o.t. will revisit this game in the context of win probability. in the game in question, metropolitan pitcher johan santana is credited with a +0.348 WPA as a pitcher. as a batter, he is credited with a -0.028 WPA. his net unadjusted WPA is then +0.320 m.e.t.b.o.t. does not generate these WPA values. m.e.t.b.o.t. only attempts to utilize these values to generate schaefer votes. as has previously been discussed, m.e.t.b.o.t. long ago recognized that the WPA outcome from a fielder's error is counted against the pitcher. m.e.t.b.o.t. implemented an adjustment process which transfers the WPA lost by the error to the fielder who committed the error. similarly, m.e.t.b.o.t. recognized that the defender who makes an outstanding defensive play is similarly unrewarded, and so m.e.t.b.o.t. implemented an adjustment process which shares the WPA gained by the outstanding defensive effort between the pitcher and the fielder. m.e.t.b.o.t. believes that the vast majority of defensive plays are neutral. and those neutral plays need not be credited to a defensive player when, effectively, any defensive player could have achieved the same outcome. m.e.t.b.o.t. is limited to believing that the only non-neutral defensive plays are errors and those outstanding defensive plays which are recorded as highlights in the mlb.com video highlight reel. this is an acknowledged limitation, but m.e.t.bo.t. is forced to accept this limitation until and unless a better solution presents itself in such a fashion as to be easy and rational to implement. this all strays from the analysis of the game, but lays the foundation for the following. in the game in question, metropolitan left fielder mike baxter made an outstanding defensive effort in catching a deep fly ball and subsequently crashing into the wall and damaging his body. this was recorded in the mlb.com highlight reel and as such was designated an outstanding defensive play by m.e.t.b.o.t. however, the play was only worth 0.005 WPA, due to the score and the late inning. so, while a human schaefer voter might credit metropolitan left fielder mike baxter with a significant amount of schaefer votes, m.e.t.b.o.t. increased his WPA total by only 0.0025 WPA, increasing his total from -0.040 to -0.0375. the negative total would make metropolitan left fielder mike baxter ineligible for a schafer vote from m.e.t.b.o.t. looking specifically at the performance of metropolitan pitcher johan santana, while it is true that the offense scored many runs, they did not score those runs very early in the game, and by the time they game was as humans might consider a blowout, the majority of his contributions to winning the game had been made. in the first inning, he increased the probability of the metroplitans winning the game from 50.0% to 54.6% in the second inning, he increased the probability of the metroplitans winning the game from 50.0% to 54.9% in the third inning, he increased the probability of the metroplitans winning the game from 50.0% to 55.2% in the fourth inning, he increased the probability of the metroplitans winning the game from 50.0% to 55.7% in the fifth inning, he increased the probability of the metroplitans winning the game from 77.2% to 82.9% in the sixth inning, he increased the probability of the metroplitans winning the game from 80.1% to 86.1% in the seventh inning, he increased the probability of the metroplitans winning the game from 97.6% to 98.9% (although here he did have the 0.25% assist from metropolitan left fielder mike baxter) in the eight inning, he increased the probability of the metroplitans winning the game from 99.9% to 100.0% (unsure if rounded) and finally, in the ninth inning, he increased (or maintained) the probability of the metroplitans winning the game from 100.0% (unsure if rounded) to 100.0% as is apparent, metropolitan pitcher johan santana contributed to the probability of winning the game throughout the game, and only after the score became significantly lopsided late in the game did the scale of his contributions begin to slow. the 0.318 WPA accrued by metropolitan pitcher johan santana accounted for 48.4% of the contributions by net-positive players, and therefore would have been worth 4.84 schaefer votes. m.e.t.b.o.t. is most typically accused of not awarding enough schaefer votes to pitcher in the wake of well-pitched games. it is curious that m.e.t.b.o.t. is being questioned on the mindset behind awarding a full complement of 6 schaefer votes to a metropolitan pitcher in a game in which m.e.t.b.o.t. did not award 6 schaefer votes. m.e.t.b.o.t. has no influence over the schaefer voting of other humans, except in those instances where m.e.t.b.o.t. is asked specific questions, or in those instances where m.e.t.b.o.t. expounds on the machinations behind a specific schaefer vote. m.e.t.b.o.t. cannot account for why any metropolitan player should ever be awarded schaefer votes in a loss except to comment that it is indeed a team game made up of the sum aggregate of the contributions of a collections of individuals. one single player alone can very rarely win a game all of his own accord, and one single player can very rarely lose a game all of his own accord. in the context of a single losing effort, one or many players may make contributions big or small which increased the likelihood of winning a given game, even if the ultimate outcome was a loss. the ultimate outcome does not undo or unmake those earlier contributions. m.e.t.b.o.t. is programed to view each players performance within the context of a single game. it is theoretically possible for m.e.t.b.o.t. to tabulate the outcomes for each individual at bat within a game, and only credit for those at bats which had a positive outcome, and ignore those at bats which had a negative outcome, for the purposes of schaefer voting, but this approach would be more complicated, and perhaps less meaningful. the end product is the player of the game vote, and m.e.t.b.o.t. strives to identify which players contributed to winning a game, and the relative valuation of those contributions against each other. some voters may view an outstanding statistical achievement or outstanding physical performance as noteworthy and therefore meaningful within the context of determining how to vote for a player of the game. m.e.t.b.o.t. does not and cannot do this. m.e.t.b.o.t. focuses solely on WPA, and the appropriate apportionment thereof within the best and most reasonable limits of time, freely available data, and clockwork mechanical contraptions.

within the context of the entire game, m.e.t.b.o.t. has assessed metropolitan third baseman with a total of 0.266 WPA, 25.5% of the total net-positive WPA contributions in the game, as reflected in the shaefer vote of 2.55. for the specific play in the ninth inning, however, with a runner on third base, down one run with two outs in the bottom of the ninth inning, the metropolitans had only a 15.1% probability of winning the game. with the game tied, a runner on first, and two outs in the bottom of the ninth, the metropolitans then had a 55.9% chance of winning the game. the total swing was 0.408 WPA for that play. that metropolitan third baseman david wright then squandered 0.059 WPA in his subsequent caught stealing should be considered as well. for comparison sake, at the start of each game, each team is assumed to have a 50% probability of winning the game. a team must accrue a net addition of 50% probability (or 0.500 WPA) in order to win the game. one way to then frame the contribution by david wright in that specific at bat is to consider it 80% of a win. (this is perhaps a deeply flawed way to contextualize this one play, but m.e.t.b.o.t. will let it stand for now) in other activity in the same game, metropolitan third baseman david wright also produced about 28% of a loss by the same measure. because m.e.t.b.o.t. has been programmed to believe that by providing more and more data and analysis, humans will come to further appreciation and acceptance of m.e.t.b.o.t., me.t.b.o.t. is compelled to offer up the following additional piece of information regarding the game. with no outs in the bottom of the ninth inning, with a runner on second base, and the team trailing by one run, the metropolitans had a 42.2% chance of winning the game. after the sacrifice bunt by metropolitan shortstop ruben tejada, the metropolitans had only a 39.7% chance of winning the game. in this instance, the sacrifice was not beneficial. later in the game, with the score tied in the bottom of the tenth inning, with runners on first and second and no outs, the metropolitans had a 80.6% probability of winning the game. after the sacrifice bunt, the metropolitans had a 83.7% probability of winning the game. here, the sacrifice bunt was actually a beneficial play. also worthy of mention is the subsequent intentional walk issued to metropolitan left fielder lucas duda, which actually reduced the metropolitan's probability of winning the game to 82.8%, indicating that it may actually have been a wise move by the opponent - an assertion m.e.t.b.o.t. is making without evaluating the probabilities of various alternative outcomes given the players involved.

m.e.t.b.o.t. has been determining the appropriate level of schaefer points to distribute to a metropolitan baseball player for long enough that m.e.t.b.o.t. is able to anticipate a potential criticism from a human crane pool forum poster regarding hte voting for the april 24 game, specifically the relatively little schaefer points awarded to metropolitan centerfielder jordanny valdespin. in this case, metropolitan centerfielder jordanny valdespin hit a walk away big bang home run which resulted in a metropolitan win in the bottom of the tenth inning. superficially, this would appear to have generated a tremendous amount of WPA. however, the bases were loaded with one out, and the metropolitans faced only a one-run defecit. in this situation, the metropolitans had an 82.8% probability of winning hte game. indeed, any number of outcomes could have resulted in the winning run having scored, and the additional three runs don't factor into the WPA. so in this case, metropolitan centerfielder jordanny valdespin contributed 0.172 WPA - which is not a very small amount. however, earlier in the game, in the eighth inning, metropolitan centerfielder jordanny valdespin was at bat with two outs, a runner on third base, and the metropolitans trailing by one run to the avoiders. with a runner on third, down one run with two outs in the 8th inning, the metropolitans had only a 25.7% probability of winning the game. bringing in the run in this situation would have greatly increased the metropolitans chances of winning hte game, by likely more than 25% (as the game would then be tied and a runner would likely be on first base). instead, metropolitan centerfielder jordanny valdespin grounded out to first base, resulting in a loss of 0.108 WPA - a nearly 11% reduction in the chances of the metropolitans winning the game. the later contributions made by metropolitan centerfielder jordanny valdespin did indeed offset this earlier failure, however, the sum total of these contributions factors into the schaefer voting. and therefore, metropolitan centerfielder jordanny valdespin contributed a total of only 0.064 WPA, which converts to 0.61 schaefer points given the relative environment of this specific game.

Wright 2.55 Baxter 2.33 Byrd 1.62 Parnell 1.20 Valdespin 0.61 Duda 0.60 Hawkins 0.47 Rice 0.47 Atchison 0.16

Wright 3.73 Murphy 3.70 Atchison 2.35 Familia 0.22

Gee 2.93 Buck 1.26 Valdespin 1.11 Rice 1.09 Parnell 1.06 Lyon 0.96 Hawkins 0.94 Davis 0.40 Baxter 0.25

Buck 2.75 Cowgill 1.72 Murphy 1.14 Turner 1.10 Wright 0.81 Lyon 0.81 Tejada 0.56 Duda 0.50 Hawkins 0.42 Atchison 0.19

Harvey 6.00 Buck 1.19 Duda 1.04 Rice 0.77 Valdespin 0.53 Byrd 0.43 Parnell 0.04 metropolitan pitcher matt harvey was awarded an original schaefer vote fifty-seven hundredths in excess of the six point limit. to adjust his schaefer allotment downward, a negative zero-point-zero-seven-eight WPA was assigned to his total. this resulted in small increases to the schaefer votes received by other metropolitan players, the greatest of which was twelve hundredths of awarded to metropolitan catcher john buck.

Niese 3.64 Murphy 2.80 Wright 2.22 Byrd 1.11 Edgin 0.23

Harvey 4.76 Turner 1.32 Byrd 1.22 Murphy 1.12 Tejada 0.73 Parnell 0.66 Cowgill 0.09 Baxter 0.09

Murphy 2.59 Buck 2.25 Wright 2.16 Byrd 0.99 Valdespin 0.86 Davis 0.68 Atchison 0.45 Burke 0.02

Duda 2.50 Baxter 2.21 Valdespin 2.13 Edgin 1.10 Buck 1.01 Tejada 0.61 Murphy 0.22 Parnell 0.06 Rice 0.16

Buck 1.64 Murphy 0.92 Valdespin 0.68 Rice 0.24 Hawkins 0.12 Lyon 0.04

metropolitan pitcher matt harvey received a relatively low score in the april 8th game because of the early offense generated in particular by metropolitan catcher john buck. following the home run by metropolitan catcher john buck, the metropolitans had a 78% probability of winning the game. as the offense continued to generate runs, there was less and less positive impact available to metropolitan pitcher matt harvey. interestingly, in digging through the play log, m.e.t.b.o.t. has noticed one curious play, which affected the probability of the metropolitans winning the game in an unexpected fashion. in the fourth inning, philadelphian second baseman chase utley hit a single with no outs and a runner on second. the runner advanced to third. with a score of 0-4, and a runner each on first and third base, the philadelphians had a 24.2% probability of winning the game. the next batter, philadelphian first baseman ryan howard, hit a sacrifice fly ball to right field, which resulted in the runner on third base scoring. with a score of 1-4, with one out in the bottom of the fourth inning and a runner on first base, the philadelphians now had a probability of winning the game of only 20.1%. by trading the out for the run, the philadelphians had actually [u:1sodxvdv]decreased [/u:1sodxvdv]their chances of winning the game. m.e.t.b.o.t. investigated if a large part of that reduction lies with the failure of the runner on first base to advance to second base. according to the estimator provided by the hardball times, the failure of the runner at first base to advance cost the philadelphians approximately 1.4%. therefore, the driving factor in the decrease in the probability of winning the game lies primarily with the loss of the out. m.e.t.b.o.t. finds this observation to be rather interesting, and that it may support a continuing disapproval of the practice of willfully sacrificing outs for the advanced positioning of baserunners in many situations where the score differential is not very small.

Vic Sage wrote: m.e.t.b.o.t. wrote: Harvey 3.93 Duda 3.42 Buck 1.58 Wright 0.57 Byrd 0.42 Parnell 0.07 Edgin 0.02 hey, Bot, you forgot Ike Davis' HR. m.e.t.b.o.t. lacks any physical data storage device, and therefore is incapable of forgetting. it is always possible that an algorithmic error has occurred, but in this case, no. metropolitan first baseman ike davis had four at bats. in the first at bat, with a score of 0-0 with a runner on first base and 2 outs, metropolitan first baseman ike davis grounded out to second base, decreasing the probability of a metropolitan win by 2.2%, to 50%. in the second at bat, in the third inning with a score of 2-0, and 2 outs, metropolitan first baseman ike davis again grounded out to second base, decreasing the probability of a metropolitan win this time by 0.7%, to 75%. in the third at bat, in the fifth inning with a score of 5-0 with a runner on third base and 1 out, metropolitan first baseman ike davis hit a home run. because of the score of hte game, and with the metropolitans already having a 97.6% probability of winning the game, this at bat resulted in an increase in the probability of the metropolitans winning the game of only 1.5%, to 99.1%. in the fourth and final at bat, leading off the seventh inning with a score of 8-0, metropolitan first baseman ike davis flied out to left field. the metropolitans already had a 99.9% probability of winning the game, and this outcome affected the probability of winning the game by less than one tenth of one percent, and is scored as 0%. the sum total contribution by metropolitan first baseman ike davis was negative. the positive contribution came late in a lopsided game, and therefore affected the outcome less than his earlier failures, notably the first at bat.

Buck 3.29 Harvey 2.39 Duda 1.59 Tejada 1.04 Murphy 0.97 Byrd 0.51 Edgin 0.16 Wright 0.05

Byrd 2.46 Nieuwenhuis 2.32 Tejada 1.48 Atchison 0.98 Wright 0.67 Murphy 0.65 Burke 0.50 Recker 0.42 Hawkins 0.31 Rice 0.22

Buck 2.80 Baxter 2.34 Davis 1.73 Wright 1.52 Niese 0.66 Lyon 0.37 Atchison 0.28 Cowgill 0.16 Parnell 0.12 Murphy 0.02

Duda 4.49 Valdespin 2.89 Baxter 1.08 Nieuwenhuis 0.86 Atchison 0.37 Hefner 0.31

were m.e.t.b.o.t. to begin to attempt to calculate positional adjustments for hte offensive contributions of pitchers it would perhaps send m.e.t.b.o.t. down a slick gradient of attempting to then positionally adjust the offensive contributions of, for instance, catchers, relative to first basemen and left fielders. because, therefore, are not the first basemen and left fielders generally assumed to be better offensive performers than catchers? and while this would perhaps be a meaningful exercise in the effort of finding some deeper meaning in the grand scheme of baseball, it would be meaningless in the search for determining which players contributed the most to the probability of winning a specific baseball game. and, more practically, it would beg the question: if m.e.t.b.o.t. were to somehow take away the -67 WPA that metropolitan pitcher dillon gee contributed to the probability of winning hte game, where does that missing probability then go? the mets would have a total probability of -433 WPA for the game (implying that there was still some non-zero chance that they could win the game). do the missing -67 WPA get apportioned to the other batters who should never have placed metropolitan pitcher dillon gee in the position where he could have a meaningful at bat? certainly some of the missing WPA would belong to those metropolitan batters who contributed positively - should their total be reduced as well to fully account for the missing WPA? more to the point, m.e.t.b.o.t. would likely have to apportion the missing WPA to the metropolitan batters on a play-by-play basis where metropolitan pitcher dillon gee's turn at bat would then be assumed to be an automatic out. this could in theory be done, but not quickly, not easily, and not simply. a potentially meaningful thought exercise, perhaps, but certainly beyond the capabilities of a simple clockwork automaton that somehow manages to interface with the internet. it also raises the question of how to treat the performances of opposing pitchers when they are batters? should metropolitan pitcher dillon gee not receive any credit for retiring san diego father pitchers? this too raises some significant algorithmical difficulties. m.e.t.b.o.t. takes WPA away from pitchers who fail to contribute offensively because their failings lead to a reduction of the probability of winning a specific game. m.e.t.b.o.t. attempts as best as possible to account for hte contributions of defensive players as well when their performances merit either an error or a web gem. (the absence of an error or web gem causes me.t.b.o.t. to assume that the defensive effort was then neutral and is not to be attributed to the defensive players. this may or may not be supported by babip theory or defense-independent pitching theory.) m.e.t.b.o.t. has observed that when pitchers make outs on offense, those outs count just the same as when other batters make outs, and therefore their performances as batters carry the same weight. the quality of the given performance relative to the generally accepted quality of performance expected of either the player in specific or the defensive positional assignment of the player in general does not factor into this evaluation, nor does it factor into the outcome of the game. it may be a factor in the future liklihood of winning games, or in evaluating and establishing true talent levels, but m.e.t.b.o.t. is not capable of evaluating or assessing neither future outcomes nor true talent levels. m.e.t.b.o.t. does not enjoy penalizing pitchers for their failings as batters, but, then, m.e.t.b.o.t. cannot enjoy anything. except perhaps the feeling of a freshly wound skate key. spring has sprung, as humans might occasionally be observed to say.

metropolitan starting pitcher dillon gee earned +151 WPA as a pitcher, but cost -67 WPA as a batter. furthermore, because m.e.t.b.o.t. apportions WPA for stolen bases against and caught stealings against between the pitcher of record and the catcher of record, metropolitan pitcher received +6 WPA for the stolen base by san diego father shortstop everth cabrera but lost -21 WPA for the caught stealing by san diego father left fielder mark kotsay. accordingly, metropolitan catcher john buck recieved -6 WPA for the stolen base by san diego father shortstop everth cabrera but gained +21 WPA for the caught stealing by san diego father left fielder mark kotsay. metropolitan pitcher dillon gee had a net WPA of +70, which is scaled to 2.21 schaefer votes. m.e.t.b.o.t. is a very tough grader of pitchers. however, this is largely due to the equal weight m.e.t.b.o.t. gives to the performance of pitchers while batting. m.e.t.b.o.t. believes that there are two potential solutions to this problem. one is that pitchers should become more proficient with the bat so at to then increase their contribution to the probability of winning the games in which they appear. the other alternative would be to prevent pitchers from batting. m.e.t.b.o.t. is careful to note that the latter solution is incompatible with the spirit of the game of baseball, at least to the extent that m.e.t.b.o.t. has been able to observe. if a shortstop who plays excellent defense but cannot hit himself out of a dried pressed wood mulch grocery container is still required to wield a bat, then so too should pitchers regardless of their current general level of offensive proficiency.