One of the more fun activities in sports is debating who the best player is at a particular position. The index can play a part in that debate, but it was never really designed to be a definitive answer to that question. Second base is one of those positions where the answer is not necessarily universal. It all depends on what you want out of a second baseman and how important level of competition is. We will begin with the index and expand our search outwards to include the various parts of the game.
Career Value
| bWAR | fWAR | WS/5 | Total | |
| Eddie Collins | 124.0 | 120.5 | 114.8 | 359.3 |
| Rogers Hornsby | 127.0 | 130.3 | 100.4 | 357.7 |
| Nap Lajoie | 107.4 | 102.2 | 99.2 | 308.8 |
| Joe Morgan | 100.6 | 98.8 | 102.4 | 301.8 |
The results here are shocking enough. Most people naturally assumed that Hornsby would top the list and I did too when I first compiled it. The difference is slight obviously, but it comes in Collins’ major advantage. He was a major player for nearly two decades. Hornsby didn’t last nearly as long, but he was legitimately an all-time great for more than a decade. However, that will come when we look peak value.
Lajoie is an interesting player because he is the first player we have that straddles the19th century and 20th century. In some cases this throws value into question because the competitive nature of the league was uneven in the early years. However, by the 1890s we can assume there was some uniformity. Most of his value came after the turn of the century, so we can logically keep him here. Lajoie becomes more intriguing when we start breaking down the individual components.
This leaves us with Morgan. Morgan obviously the only player from the modern game and as such played in a multi-chrome environment. This is where terms like greatest and most dominant can be conflated. Morgan was clearly not more dominant than the other three but given arguments about the quality of play we could claim he was the greatest. Then, those arguments are based on certain assumptions too. Let’s move onto peak value and quickly put the index part of the argument to bed.
Peak Value
| bWAR | fWAR | WS/5 | Total | Index | |
| Rogers Hornsby | 93.1 | 96.2 | 72.4 | 261.7 | 619.4 |
| Eddie Collins | 78.3 | 77.2 | 75.4 | 230.9 | 590.2 |
| Nap Lajoie | 75.0 | 72.6 | 64.2 | 211.8 | 520.6 |
| Joe Morgan | 77.2 | 66.5 | 63.0 | 206.7 | 508.5 |
One of the notions of the index is the notion of separation. It is designed to separate players as much as possible for Hall of Fame fitness analysis. You take career value and peak value for instance. One is part of the other, but when you put it together it often separates players purposely. The same is true when we use three different statistical sources. Occasionally, we get some honest disagreement and that is where it pays to shine a light and see what is going on.
Eddie Collins had more win shares for his career and during his peak than Hornsby. That begs the question why. WAR exists in a runs environment. Wins are parceled out on a ten-run basis. The idea is that performance occurs in the aggregate. Over a long enough time period all of the hiccups of clutch performance will even themselves. At least that is the idea behind. The creators will tell you that it isn’t necessarily meant to be leant on over a year to year basis. It becomes more accurate the further away you stand.
Win shares are parceled out based on wins. This creates a bit of an issue itself. Players are rewarded based on their placement on a good team. In an aggregate environment, when individual performance improves team performance improves. Unfortunately, it isn’t an exact ratio though. So, it is possible that some players benefit from collective performance or suffer from a lack of support. Yet, it is also possible that when a player performs is more important than simply leaning on the aggregate. Proven clutch performers deserve credit for performing in those situations and WAR simply doesn’t do that. Does Collins benefit simply by being on better teams or he was he honestly a better performer when it mattered? That remains to be seen.
Offensive Numbers
When we distill out counting numbers we often see a very clear picture of who was the more valuable offensive player. The trouble is that we are really only considering what occurred in the batter’s box. Historians, statisticians, and scouts all agree that what occurs in the batter’s box trumps everything else most of the time, but we can obviously debate to the percentages involved. Here, we see mostly how each player performed against the league with the exception of bases per out. We will cover that down the line.
| wOBA | wRC+ | OW% | BPO | |
| Rogers Hornsby | .459 | 173 | .815 | 1.049 |
| Eddie Collins | .409 | 144 | .715 | .796 |
| Nap Lajoie | .401 | 144 | .758 | .777 |
| Joe Morgan | .372 | 135 | .702 | .818 |
To give us some idea we will start with offensive winning percentage. In a 162 game schedule, a team made of these players would win at least 113 games a season. The 2001 Mariners set the modern record with 116 wins in 2001. So, these players in their prime were better than the 2001 Mariners even with average pitching and fielding. A team made up of nine Rogers Hornsby’s would win 132 games on average. That is patently ridiculous.
Quite frankly. Hornsby is in the conversation for best right-handed hitter in the history of the game. It really isn’t fair to compare the other players to him, but they do compare favorably to each other. Morgan comes a little short, but we have to keep in mind that it is more difficult to separate yourself in the modern game. Expansion watered down the league and mitigated the effects of deviation. In other words, the standard deviation would be lower, so he might have been better.
Bases per out is a hybrid number meant to show how valuable a hitter was. It unfortunately is not normed like the other numbers, so the raw scores can be misinterpreted. Lajoie and Collins played in the Dead Ball Era when offensive numbers were depressed overall. So, seeing Morgan end up a little higher can be somewhat deceiving. Still, this is just one more category where Hornsby dominates. Naturally, this would lead most to just assume that he was the best of all-time, but there is more to the game than what happens in the batter’s box.
Fielding Numbers
| BR | FG | DWS | WSGG | |
| Nap Lajoie | +83 | +79 | 85.7 | 4 |
| Rogers Hornsby | +54 | +30 | 59.6 | 1 |
| Eddie Collins | +35 | +40 | 107.5 | 8 |
| Joe Morgan | -48 | -43 | 90.8 | 0 |
Yeah, but but but Joe Morgan won five Gold Gloves! Again, we see the folly of relying on Rawlings to tell who us who the most valuable fielder was. The system is rife with bias first of all, but more importantly only one of our players was eligible for the awards. In the past, we have looked at total zone awards and Fielding Bible awards for the modern players, but here will have to stick win shares on an individual year in and year out basis.
Based on his finish here, we can safely eliminate Morgan from the discussion of who would be the most valuable second baseman of all-time. Of course, the discussion on the greatest second baseman rages on. Morgan combined power, speed, and patience like no other player at the position. Unfortunately, he wasn’t quite dominant enough with the bat or the glove.
Of course, win shares are indicative of value over time. Morgan is second all-time in defensive innings at second base, so it makes perfect sense for him to be second in defensive win shares. The win share Gold Gloves are more indicative of defensive value from year to year. Hornsby had some defensive flexibility, but he was not valuable at any position he played according to win shares. The baseball-reference and Fangraphs metrics said he was better than the average second baseman.
This is the primary reason why we do not compile fielding statistics like we do with the index. Fangraphs and baseball-reference compare players with the average at the position while win shares compiles from the replacement level. When you have two different frames of reference you end up having two different results. The take away is that Eddie Collins was likely the best of the four and the most valuable over time. Lajoie was second in terms of greatness but had fewer innings than the others.
While it would seem prudent to eliminate Morgan at this point, those that watched him play would harp on his baserunning as an advantage. One of the peculiarities of scouting is that scouts often salivate over the so-called five tool player. Since fielding and throwing go hand in hand, that often means hitting, hitting for power, and running. The trouble is that they assume every tool is equal. That being said, we should look at baserunning before we write off Morgan.
Baserunning Statistics
| BsR | Rbaser | |
| Joe Morgan | 79.0 | 80.0 |
| Eddie Collins | 42.3 | 40.0 |
| Rogers Hornsby | -1.8 | -9.0 |
| Nap Lajoie | -3.0 | -11.0 |
There is a distinct difference ordinal and interval data. Your ranking of these second baseman largely depends on your adherence to either ordinal or interval data. Ordinal data simply ranks each player in order. In that case, Eddie Collins would be your man. He ranks either first or second in every category individually. So, while Hornsby was the best hitter, he was clearly deficient in fielding and running.
Morgan becomes viable when you consider each facet of the game individually, but when you consider it altogether you get to interval data. That asks how much better someone is than someone else. This is where we discover that hitting is more valuable than fielding which is more valuable than baserunning. So, we can salivate over the five tools, but we need to keep it all in perspective.
Rogers Hornsby’s advantage with the bat is probably enough to carry him when we add in fielding and baserunning. However, the finish is a lot closer than it would appear. In a qualitative way, people might prefer a more well-rounded player and Eddie Collins is the most well-rounded second baseman.
halloffameindex 