When your first thought upon starting something is “I probably can’t carry this through to the end, but whatever,” you should probably start doing that thing.

My best friend from high school and I walked across our hometown of Springfield for 2 1/2 hours once, just because we could see the capitol building from the Buffalo Wild Wings we were at and thought, “I wonder if we can walk there. Probably not.”

In my excitement for the NBA season’s beginning, I decided to start doing a tally that would turn my feelings on the MVP race into a simple statistical exercise. I did this mostly because I like to discredit LeBron James. And I knew I’d probably find a way to make a case for someone else down the line, and wanted to check myself before I wrecked myself.

My philosophy on the NBA MVP ballot is simple: The player who is the best player on the winning team in the most games should be the MVP. I just had to come up with a method to support my philosophy. I decided to jot down who I thought was the best player on the winning team from every game. I could have picked 10 guys, but I decided to expand my idea to every game in the entire league, and to jot down the losing team’s best player as well, to see what this would show me.

This prompted a few questions. How many hundreds of games are there in the NBA per season? How much time do I have on my hands? How much do I know how to use an Excel spreadsheet? The answers were 1,230, enough, and not at all, respectively. Until Christmas, I wrote down every game’s winning and losing teams’ best players in a text document. An engineering friend, Jon, kindly set up a spreadsheet that was much easier to use and saved me the trouble of calculating percentages, which got a heck of a lot harder after 10 games.

To analyze the game’s best player, I used the very subjective but reasonably effective method of looking at a game’s box score and deducing who on each team was worthy of the most credit in a win or put forth the best effort in a loss.

The results were fascinating.

What the results can show

It’s hard to justify calling my subjective thoughts a “statistic,” but that’s the most appropriate term I could think of (though you may think of it as an “arbitrary number” if you wish, jerk). I’ve come to calling the statistics “leads,” for when a player leads his team, “wins,” for when a player leads his team to a win, and “losses,” for what a losing team’s best player is credited with. “Losses” is the most inappropriate term here, and I’ll get to why in a bit.

But leads and wins can show us a lot, not only about how dominant one player is compared with another player, but how different team dynamics work. Do good teams have one or two leaders? Or is it better with several? Do those deadly duos like LaMarcus Aldridge and Damian Lillard share the workload evenly or does one take on more responsibility? And finally, who the hell leads a 15-win team like the Milwaukee Bucks on a nightly basis?

I really found out about all of them. Or rather, I suppose, found out how I feel about all of them. I certainly gained insights I wasn’t expecting to gain.

Oh, and I have numbers that directly indicates Kevin Durant should be MVP over LeBron James. So that’s pretty cool.

What the results can’t show

Obviously, leads are a limited resource. There are 2,460 leads handed out in a given year. Sometimes the talent level of the league as a whole fluctuates. The amount of leads given out would not. The Heat get as many leads as the Bucks, but I don’t think anyone on the Bucks is more valuable to a team than any Miami Heat starter (except maybe Mario Chalmers).

So what this doesn’t tell us is who the top 100 players in the league are. At a certain point, you have to understand that Chris Paul’s numbers do take away from Blake Griffin’s MVP case, but not as much as they do in this format. I think ultimately, the top leaders are worthy of being there, but once you go past the top 5 or 10 players, you run into trouble.

More ambiguous is just what we can take from a player “losing” a game. Obviously one has to play well in order to earn a lead. Is a player who earns more than 30 losses but also more than 30 wins more valuable than a player who loses less than 15 games but barely wins 20? It’s a tough comparison — would you take Carmelo Anthony over LaMarcus Aldridge?

But certainly from this we can’t judge who is a “worse” player, the ideas of losses and leads directly counteract each other. We also can’t judge who is the least valuable. You’d be better off combining offensive efficiency and usage rate, and evaluating that.

Missing components I wish I added

I wasn’t completely comprehensive. I could have determined not only who was each team’s best player, but who was second and third best, all the way down to seventh or eighth. That way when a player really takes a night off, and goes from being first best to fifth or sixth, it’s documented and can hurt their MVP resume. In a lot of ways, it looks better to avoid the loss for the sake of keeping a high win percentage (percentage of a player’s leads his team wins), but that means a player looks better for playing not as well, which is backwards.

We could also see, then, how a team fares when a given player is its second best player. That would clarify the record of someone who loses and wins a lot of games — maybe it’s dependent on who’s complementing their production, not their production in and of itself.

I’d have kept more accessible records regarding home record vs. away record, and dates of wins and losses and leads, to determine things like who’s been the MVP since the All-Star Break? Or since Jan. 1? Examining smaller samplings of games can reveal a lot about the way a team’s season has gone.

I’d also keep track of players’ positions. It seemed a lot of times big men would have good games against each other, or point guards. That correlation would be worth exploring.

There’s an inherent inconsistency to this tally, with me being a human and all. If I was using a formula (which, primitive as it sounds, I would often add up total points, rebounds and assists to get a better idea of who contributed what), I could guarantee consistency. But given a choice to go strictly by a formula or strictly by my opinion, I’d prefer to keep the human element involved. Because humans play basketball and humans decide the MVP winner, which is, after all, the goal of this project.

I didn’t factor in games missed by individual players. This was another element that made sense within the confines of my narrow starting goal, but failed me within the broader context I began to delve into as the season went on. As of now, the impact of given players who missed time is understated.

Lastly, it’s unfortunate to only have one year’s worth of statistics in any category. It could be tremendously helpful to see whether this logic would have crowned James MVP last year (undoubtedly it would have) or Derrick Rose in 2011 (it may have). Also it would be useful as a tool to compare players’ regression or progress.

“Wins” leaders (number of games a player led his team in a win)

1. Kevin Durant, 48

2. LeBron James, 42

3. Carmelo Anthony, 31

4. Kevin Love, 30

5. Blake Griffin, 27

T6. Paul George, 26

T6. Al Jefferson, 26

8. LaMarcus Aldridge, 25

T9. James Harden, 24

T9. Zach Randolph, 24

T9. Stephen Curry, 24

And here’s your MVP award. Kevin Durant was simply the best player for the winning team the most times this season. I found it funny that the day LeBron essentially conceded the MVP to Durant, saying he’s been the “most consistent basketball player as far as the MVP,” was the very day after Durant mathematically eliminated James from leading in the wins category.

Interestingly, no one has really made a case for Carmelo Anthony to even be on the ballot because of how poorly the Knicks have played, yet Kevin Love gets tossed in as a top-5 MVP finisher despite the Wolves finishing 10th in their conference and underachieving preseason expectations (not to mention underachieving their own output — how do you finish below .500 with a +2.7 average point differential?). I think Anthony is very worthy of a top-5 MVP finish, given how consistently he put in the effort throughout the year. The Knicks only had 37 wins, and Anthony pushed them to 31 of them. That’s darn impressive.

The only other surprise on this top 11 list would probably be Zach Randolph. He was under-credited for keeping the Grizzlies afloat while Marc Gasol was out with injury. When Gasol came back, he remained the team’s predominant motor as they moved up the standings and eventually into the seven seed. It’ll be interesting to see if they do damage deep into the playoffs.

“Leads” leaders (number of games a player led his team, win or loss)

T1. Kevin Durant, 65

T1. Carmelo Anthony, 65

3. LeBron James, 61

4. Kevin Love, 57

T5. Al Jefferson, 46

T5. DeMarcus Cousins, 46

7. Stephen Curry, 45

8. Paul George, 43

9. John Wall, 42

10. Blake Griffin, 40

This really takes an unbiased look at who the league’s stat stuffers are, without providing a whole lot else. Anthony’s total is more ridiculous when you consider that he missed five games, of which he probably would have claimed four. Durant only missed one game, which deservedly strengthens his grip on the MVP trophy. James missed five games, though two were to rest up for the postseason.

DeMarcus Cousins is in good company on this list, and I think it’s time we begin to view him as an elite player. Just because he’s such a grump doesn’t mean he isn’t talented. The Kings need to do their best to accommodate him, or next year he’ll lead the next category.

“Losses” leaders (number of games a player led his team in losses)

1. Carmelo Anthony, 34

2. DeMarcus Cousins, 30

3. Kevin Love, 27

4. Brandon Knight, 23

5. Stephen Curry, 21

T6. Paul Millsap, 20

T6. Pau Gasol, 20

T6. Al Jefferson, 20

T9. LeBron James, 19

T9. John Wall, 19

T9. Anthony Davis, 19

T9. Kyrie Irving, 19

T9. Enes Kanter, 19

Here’s where it gets weird. In case you were trying to dream up a category in which LeBron James ties with Utah Jazz off-and-on-starting center Enes Kanter, well, here you go. Here we have an amalgamation of bad situations and problematic team dynamics. Sometimes it’s lack of a supporting cast. Other times it’s being a stat hunter who doesn’t make teammates better. And sometimes it’s just because a team has a leader and loses and wins with him (Durant has 17 losses, so he’s not far off this list).

With so many differing reasons for being on here (for instance, James has 42 wins, Kanter has 1, yet they’re tied), it’s hard to tell what being on this list really means. The only thing I’d say it means is that something’s not right and the team’s front office needs to figure it out.

For me, the red flags go off at the sight of Stephen Curry’s name. He’s often considered a candidate to be an MVP also-ran, but his team actually plays better when he’s not its central focus. That’s scary. I’d probably say it has to do with the Warriors being a shooting team, and Curry taking the hit for when his teammates go cold. But here’s where I’d like to see who a team’s second and third best players are. Does Curry need Klay Thompson to succeed for the Warriors to win (Thompson had five leads and five wins, by the way)? Or would someone like David Lee or Andre Iguodala provide a better complement to Curry’s lights out shooting?

This category seems to raise more questions than answers. 

Eliot is a senior in Media. He can be reached at [email protected]. Follow him on Twitter @EliotTweet.