I know that this question has probably been asked hundreds of times, but I have done some research and work and would like to know your opinion of a point system that I have come up with. It is not like a typical point system, rewarding much for singles, doubles and triples, but instead runs on the basis of the reason batting and pitching exist: to get on base and score runs, or disallow such. I know that point systems generally are considered "unfair," but I feel compelled to use one, because it seems like it would be exciting. Hopefully this system can improve a little.
The system I have come up with deducts points per each at bat (.3), but gives 1 point for each hit. The idea here is that you will be only making points (provided you only get the point for the hit) if you bat at around a .300 average. However, it is easier to make more points as doubles, triples and runs are still rewarded, but on a lesser scale:
Basically, the Batting statistics are made to reflect the goals of getting on base and scoring. The gimmick of stolen bases is still there, but I decided to put that in there just to add some "variety" to the scoring. The categories marked with an asterisk are the ones I am most keen on changing: strikeouts I see as somewhat gimmicky, but I could go with a -0.2 point deduction. RBI's, I believe, are important, but I feel that them being worth more than a double's bonus might be much. I also intend on putting two "bonus" categories in: Grand slam, probably worth 0.5 points, and hitting the cycle, most likely worth 3 or 4 points.
Pitching follows the same philosophy: don't allow hits or runs.
The point here is to only allow points with, at the very least, an 8 ERA not including points lost from hits. I originally planned to set ER at -1.2. Your advice on weather to switch that back. Strikeouts are worth something, but not overly much; there are many other (maybe preferred) ways of getting batters out other than strikeouts.
"Bonuses" here are complete games and shutouts. Saves are 2 points in order to encourage the use of closers, and holds are 1.2 to do the same.
I have created an Excel spreadsheet with lots of graphs and stats on it. It includes the full used states from the top 40 batters and pitchers last season. Scatter plot graphs are included for points vs WAR, AVG, ERA, Runs created, and many more stats. They show relatively good correlation, especially with runs created in batting and WAR in pitching. I'll post the URL when I can, the fourm won't let me right now.
Any comments/advice is appreciated. Thanks!
Last edited by gonats03 on Wed Jan 23, 2013 9:58 pm, edited 2 times in total.
Here are the URL's of the spreadsheet. to get there, just remove all of the (+) symbols. Feel free to view it (URL 1), or edit it (excel document, URL 2)
I've actually been looking for a good points system, and I've tested this one in my league scoring. First thing I noticed is that it's extremely balanced, the top hitters and pitchers all accumulated about the same amount of points, and through the top 30 of last year for each hitters and pitchers, it's well balanced. My only gripe is the scoring would be low (realistic, but low), I guess baseball scoring is low but in my leagues I like a lot of points and I messed around with multiplying each stat by 4 (to leave the same balance, but just increase points).
This worked as well, instead of the top batter getting 170 season points, he got about 700 (so instead of say under 10 pts per week he would get 30+ per week)... I guess it all depends on your preference. I do like this system alot, there's obviously a lot went into this system, I'm still trying to wrap my head around weekly matchups - say for pitchers who go two starts, maybe they total 15 innings = that's alot of points per inning (if your scoring 1 per or 4 per if you increased it), but the balance with runs/hits/walks works out... maybe i'll try slimming it down to multiply by 3.
Looking at Verlanders first week last year, with 2 starts = he would get 12 points (off 16 innings, 6 hits, 4 er, 3 bb and 14 k's). Looking at Cabrera's first week, he would get 15 points (off 33 abs, 7 runs, 3 homers, 9 rbi, 6 bb, 5 so) They are both pretty good weeks, I know its a small sample but I'm just looking to compare this system.
Chaos151p wrote:I've actually been looking for a good points system, and I've tested this one in my league scoring. First thing I noticed is that it's extremely balanced, the top hitters and pitchers all accumulated about the same amount of points, and through the top 30 of last year for each hitters and pitchers, it's well balanced. My only gripe is the scoring would be low (realistic, but low), I guess baseball scoring is low but in my leagues I like a lot of points and I messed around with multiplying each stat by 4 (to leave the same balance, but just increase points).
This worked as well, instead of the top batter getting 170 season points, he got about 700 (so instead of say under 10 pts per week he would get 30+ per week)... I guess it all depends on your preference. I do like this system alot, there's obviously a lot went into this system, I'm still trying to wrap my head around weekly matchups - say for pitchers who go two starts, maybe they total 15 innings = that's alot of points per inning (if your scoring 1 per or 4 per if you increased it), but the balance with runs/hits/walks works out... maybe i'll try slimming it down to multiply by 3.
Looking at Verlanders first week last year, with 2 starts = he would get 12 points (off 16 innings, 6 hits, 4 er, 3 bb and 14 k's). Looking at Cabrera's first week, he would get 15 points (off 33 abs, 7 runs, 3 homers, 9 rbi, 6 bb, 5 so) They are both pretty good weeks, I know its a small sample but I'm just looking to compare this system.
Thanks a lot for commenting! the first thing I did was made an "outline" of the pitching points and batting points, then tweaked them to make them close. Between the top 40 batters and pitchers, Pitchers got 4457.7 points and batters Batters 4540.4 points. That's a difference of around 2%. It's pretty close, but probably not perfect. By adding -0.1 points to make the ERA -1.2 and adding -0.1 points to strikeouts (for batters) to make them -0.2, you get a little closer sample, with a difference of around 20. I just thought that 0.2 for strikeouts was a little much; just because I feel that it might be a bad way to get out, but I don't think it's as bad as a double play, and a bad hit is a bad hit.
The other part I'm trying to do is figure out weather the stats match real player value. This is nearly impossible to determine, but I guess it would mean the greatest correlation between the most stats. If you look at the graphs, pitching correlates relatively well with WAR, WHIP and ERA, which I'm happy about. Batting doesn't correlate with AVG or WAR as much, but correlates really well with Runs Created and OPS. Naturally, power hitters are going to be rewarded (home runs, triples and doubles), but hopefully won't be rewarded tons more, hopefully making a good leadoff hitter worth a good amount because of the amount hits are worth vs the rest of the lot.
That's a good idea to set a multiplier for the points, as it would make it a bit more high scoring. I'll probably end up multiplying them by 3 or 4 to make the scoring a bit higher. Just trying to get the basic outline down. Thanks for the reply. I'll try to do a little more stat-crunching in the meantime.
I've also compared a few hitters with this decimal system (multiplied by 3) to a standard points scoring system (1B = 1, HR = 4, etc). Three batters, Posey/Rollins/Hamilton
Posey had 78 runs, 24 homers, 103 rbi (with a .333 avg) = about 446 points (standard) and 433 (decimal). Which ranked him 21st among hitters in the standard scoring, and 7th overall in the decimal scoring. He wound up with more points for hits then taking away for AB's (534 to 477) which relates to his batting average in real life. He also had 39 doubles and 69 walks to increase his value.
Rollins has 102 runs, 23 HR, 68 rbi...plus 30 steals. (with a .250 avg) = about 461 points (standard) and 287 (decimal). This is a big drop off, Rollins ranked 18th in standard but fell to 40th in decimal scoring. Most likely due to his poor average and low rbi - he did have 33 doubles and 60 walks.
In summary of these first two... Rollins ranked higher than Posey in Standard scoring, even though his avg. was .250 compared to Posey's .333 - they had about the same amount of runs/rbi total... but Rollins sneaks in the 30 steals. So in standard scoring, 30 steals weighed more then 80 points in average... that's unrealistic.
Just to prove Rollins should in fact be ranked near the 40th instead of inside the top 20... here's Hamilton's numbers - and Rollins and Hamilton ranked close in standard. So by this comparison...
Hamilton had 103 runs, 43 HR, 128 RBI.... with .250 average (7 steals and 31 doubles) = he ranked 15th in standard and 10th in decimal. But in the standard league, Rollins was right under Hamilton's = 230 total runs/rbi and 40+ homeruns
Rollins had a decent year, but he should be ranked near Pagan (95 runs, 8 homers, 29 steals, .287 avg) not Hamilton (103 runs, 43 HR, 128 RBI)
Chaos151p wrote:I've also compared a few hitters with this decimal system (multiplied by 3) to a standard points scoring system (1B = 1, HR = 4, etc). Three batters, Posey/Rollins/Hamilton
Posey had 78 runs, 24 homers, 103 rbi (with a .333 avg) = about 446 points (standard) and 433 (decimal). Which ranked him 21st among hitters in the standard scoring, and 7th overall in the decimal scoring. He wound up with more points for hits then taking away for AB's (534 to 477) which relates to his batting average in real life. He also had 39 doubles and 69 walks to increase his value.
Rollins has 102 runs, 23 HR, 68 rbi...plus 30 steals. (with a .250 avg) = about 461 points (standard) and 287 (decimal). This is a big drop off, Rollins ranked 18th in standard but fell to 40th in decimal scoring. Most likely due to his poor average and low rbi - he did have 33 doubles and 60 walks.
In summary of these first two... Rollins ranked higher than Posey in Standard scoring, even though his avg. was .250 compared to Posey's .333 - they had about the same amount of runs/rbi total... but Rollins sneaks in the 30 steals. So in standard scoring, 30 steals weighed more then 80 points in average... that's unrealistic.
Just to prove Rollins should in fact be ranked near the 40th instead of inside the top 20... here's Hamilton's numbers - and Rollins and Hamilton ranked close in standard. So by this comparison...
Hamilton had 103 runs, 43 HR, 128 RBI.... with .250 average (7 steals and 31 doubles) = he ranked 15th in standard and 10th in decimal. But in the standard league, Rollins was right under Hamilton's = 230 total runs/rbi and 40+ homeruns
Rollins had a decent year, but he should be ranked near Pagan (95 runs, 8 homers, 29 steals, .287 avg) not Hamilton (103 runs, 43 HR, 128 RBI)
That's a good visualization of the points system, and one that I like. I'm liking the way it's coming out as, and I hope it will look better.
One thing that I'm probably going to change is the bonus for saves. I overlooked this one and it looks like top closers (with more than 50 saves) can outscore top starting pitchers. Not really the best idea. Because of this, I think it would be better to put around 1 point for each save and less for a hold.
The one thing I really dislike about the standard scoring system was the over-rewarding for stolen bases, doubles and triples, and the leaving out of AB and hits. I also don't think that starting pitchers should be awarded for QS, Wins or losses just because it's something that is unneeded and out of their control, which is something I tried to incorporate here. I'll see if I can tweak this a little more. Thanks for the reply, all the comments are really helpful.
I can't add much to this, except that I find this to be really helpful. After playing in points-based leagues for about eight years now I was just about read to throw in the towel and try a rotisserie scoring setup. I've always wanted to see a points-based scoring system that reflected actual gameplay in a more realistic way. Keep up the great work...my league may adopt this for 2013!
Hi. If you are planning on inviting managers into your league, I'd be very interested (keeper or re-draft). I used to manage a league with a similar to this scoring system. The managers' only grip was that scoring was too low, so I multiplied by 3. looks great!
Been playing around with the system again. Over the last few days, I compiled pretty much every stat imaginable from around the top 140 batters over the last year (by batting average). I decided to put every one of then in the scoring system. The top player was still Miggy, but I decided to check the correlations between the points and each stat. So I calculated the correlations for each stat using excel (there were around 220 stats), and the #1 stat that correlated the most with the points system was wRC (weighted runs created). The stat actually uses similar point-style calculations to the point system. It correlated with the points around 97%. Not far behind was wRA and wOBA. I did a scatter plot with wRC to see the similarities, and it is obvious. I think I'm gonna do this for pitching too, and then play around with it to see if I can improve on it.
Sure, I think there's room! I haven't gotten a full team together with this yet, just preparing so we can get started early. I'll send you more info as I know more about the team.
I would post this Excel spreadsheet but I can't upload files right now and uploading the file to the internet makes the formatting go awry. I'll see what I can do in the meantime.
OK, so I think that the point scoring system has been discussed, now the question of the roster is coming. The default roster in ESPN is 13 batters and 9 pitchers. Because pitchers and batters score approximately equal points, the number here should be around even, and 13/9 seems OK. The next is GP limit for pitchers in a match up. My idea is to not have this- this may mean that someone who has a ton of pitching starts one week will probably win- but then probably loose the next. Your ideas are appreciated. ESPN Fantasy opened today so I'll be doing a lot of tweaking. Any comments on the points system or the roster are helpful. Thanks!