From time to time, I’ll hear reference to BeerGraphs as if we’re trying to legislate which beers are actually the best with our numbers. I don’t really think that’s the aim — we research beer trends and report what the numbers say, and we try to offer different ways to slice available numbers so that you can make more informed decisions. There’s no way to analytically prove which beer is the best.
There’s no way to analytically prove which settings are the best for your fantasy league. It’s fantasy, it’s not real, and so whatever settings you decide are just the rules for the game you are about to play. Different settings do beget different styles of analysis, though, so it’s worth a little thinking to get it right, especially if you are about to start a league that you hope will be around a while.
I’ve got 15 leagues. I’ll just write about the pros and cons of each. Maybe that’ll help.
5×5 Old-School Roto
(BA/R/RBI/HR/SB and K/ERA/WHIP/Saves/Wins)
I’ll admit something. 5×5 is my favorite. It’s just the classic game. We know that games aren’t decided by these stats, but that’s okay. When fantasy baseball was born, these were the scoring categories. Saves and steals may become the bane of your existence — they’re flighty, require stringent waiver work, and have little to do with the outcomes of real-life games. But this is the game we decided to play. If you can’t all be at the computer as often as each other, throw a waiver period on there. Give everyone $100 Free Agency Auction Budget (FAAB), and then they have to decide how much they want to spend on that new closer, and everyone who logs in on that day/week gets a shot at the player. But if you look at these five categories, you’ll notice some balance. Power players dominate two offensive categories while speedier on-base guys have a chance to help in three. Starting pitchers are needed for strikeouts and wins, but you still have to have good relievers to help keep the ratios down and get saves.
4×4 Old-School Roto
(BA/RBI/HR/SB and ERA/WHIP/Saves/Wins)
Some think this is the classic game, but you have to be specific about which eight categories you’re talking about. The very first roto games didn’t have runs or strikeouts, but that’s just weird. A batting average / RBI / HR / SB setup weights power over on-base skills to the point that speed guys are almost an after-thought. And taking strikeouts out of the game, at this point, seems crazy. They’re huge. You could put together a great team without any front-line starters — all relievers and some ground-ball guys, and your team is lights out. Eh.
4×4 Saber Roto
(OBP/SLG/HR/R and ERA/WHIP/HRper9/K)
There’s a new 4×4 in ottoneu which might appeal to those that hate saves and steals. Your relievers can still help keep the homer rate and the ratios down, but they don’t get a counting stat that only they can provide. That is kind of nice, that you aren’t forced to chase around the kicker of fantasy baseball, the one-trick pony, just for saves. (And your worm-burner starters and stingy homer relievers get a bit of a boost, too.) Here, you’re trying to get good pitchers, regardless of role. And on the offensive side, you weight power a little bit more, but you also value speedy on-base guys (OBP and R) without putting any emphasis on a stat like steals, which are less valuable in the real-life game.
Sabermetric Points Roto
(wOBA, FIP-based points)
There’s a benefit to using points systems. You can correctly weight the events so that they line up perfectly with real-life value. That’s a lofty goal. However, when it comes to game play, I have a complaint with points leagues. Trading can be strange. It’s like in fantasy football, when you’re offering a guy with x points for a guy with y. If y>x, the person on the other end of the offer is quick to dismiss. You have to spend a lot of time convincing them that the player fits some other need, or will be more valuable in the future, blah blah blah. I just don’t like one number attached to a player’s value in fantasy sports, it’s much more fun to try and shift available resources in my mind — “I’ve got too many steals, you need steals, let me trade this leadoff guy for a middle of the order bat and solve both our problems.”
6×6 Slightly Saber Roto
(BA/OBP/R/RBI/HR/SB and K/ERA/WHIP/Saves/Wins/Holds)
There will be some here that hate these settings. They’re kind of stuck between old- and new-school. Still have the 5×5 in there, but you’ve added on-base percentage because it’s a valuable real-life skill, and you’ve added holds because… Well listen, I get it, but my longest-term dynasty league has these settings, so I’ll defend holds for a second. Holds make relievers other than closers more valuable. Perhaps a saves-plus-holds category would be better, so that it’s still just one counting stat, but making setup men a valuable part of fantasy baseball is fun. They obviously matter in real-life baseball, and it’s a whole class of player that can now be a part of the game. Other ways to accomplish this goal are to either add saves and holds together in one category if your provider allows it… or to up the number of Relief Pitcher slots in your league. If you have five reliever slots in a 12-team league, you’ll be rostering some non-closers to help with rates and ratios, and to take advantage of every starting slot you’ve got. The problem with having both batting average and on-base percentage in your league settings is that one stat is in the other. A high average guy can now kill it in two categories instead of just one.
5×5 New-School Roto
(OBP/R/RBI/HR/SB and K/ERA/WHIP/Saves/Wins)
TOUT is playing footsie with this setting, and it’s easy to see why. With one little flip of the switch, we’re now valuing the better real-life skill — getting on base as opposed to making contact and doing well on balls in play. The only thing that’s a bit weird about this is that it doesn’t go very far. It’s hard to call this new-school, even, but it is to the old versions of the game. OBP! Oh my!
Anti-Streaming Head to Head
(—— and Kper9/ERA/WHIP/Wins/Saves) or an innings cap, moves cap, FAAB strategy
One of the things that’s great about head-to-head settings is that it can be more lively. The setup allows for more trash-talking, for one. And if you’re out of it, you can right your team quicker than if you’re trying to push the roto category rock back up the hill. But an un-intended consequence for some is that the guy in front of the computer 24/7 can add a lot of value to his team by pouncing on the wire for saves, and then perhaps streaming the last few starting pitcher slots on his team. In a world where you’re just valuing strikeouts as a category, he’d be looking to win K, W and Svs every week with a reliever-heavy streaming category. That’s viable — his three categories to your two.
You could put a cap on moves to rein in streaming if you don’t like it. But then everyone, streamer or non-streamer alike, is watching their moves. It tamps down activity, and if you set it too high, you might still allow a player to stream a guy every week and take advantage of two-start pitchers. An FAAB strategy has many of the same flaws, but at least there’s a bit more strategy — go for bulk pickups at lower costs, or get the premier free agents with big bids. If you allow $0 moves in FAAB — most systems do — then someone can still stream, they just lose waiver position every time they make a $0 move. They can still stream, but there’s a penalty now.
The easiest way to make streaming less attractive, though, is to replace strikeouts with a rate stat. Now all those streamers are bringing down the K/9 category and it’s suddenly a losing strategy. In my experience, category changes are the easiest way to affect change — any other attempt to codify anti-streaming policies have unintended consequences (think the lower level of activity that happens when you put a move limit in place).
The Other Stats
(Batter: ISO, BB%, Ks, NetSB, Grand Slams; Pitcher: K/9, BB/9, HR/9, GB%, CG)
There’s some value to be gained by making your stats in your league more sabermetrically inclined, I understand the impetus. For example, why not move stolen bases to net stolen bases? If stolen bases are a positive, being caught stealing should be a negative. And by adding ground ball rate OR homers per nine, you can reward pitchers that keep the ball on the ground, which we know is a valuable skill. I don’t mind the idea. Just try to look at your stats and see, on a very basic level, which sorts of players your system values. Because if you do something like OPS, and then also add in batting average, you’re having the same problem my AVG/OBP league had. If you add GB% AND HR/9, you might be over-valuing worm-burners. Be careful, because every category you add dilutes the importance of any one category, and yet also gives you the opportunity to double-count some more and over-value a certain type of player.
In other words, say out loud what your problem with your current setting is. Get the league to say what they think. If it’s “we think steals guys are too heavily favored,” change SB to Net-SB, or take it out and replace it with OBP or something that helps top-of-the-lineup guys whether or not they steal bases. You could even use strikeouts by the batter here, since strikeouts are correlated with power. If it’s “we hate closers,” maybe consider dropping saves and adding a strikeout rate stat. It’s not that weird to have Ks AND K/9 — it just means you need to accrue strikeouts at a good rate. And you’ll have a reason to have good relievers on your team with or without the chance at getting saves. If you hate wins, sure, use quality starts. They aren’t the best thing in the world, but they’re slightly better than wins!
And don’t forget that there are things you can change other than the categories. The way you run waivers is important to the activity in your league. Your innings and plate appearance minimums and maximums change things for streamers, too. I agree with Paul Singman when he argues for lots of DL slots and long benches, but that can thin out the waiver wire, too.
One thing. Don’t use the rare stats. Especially in head-to-head leagues. The single worst settings I’ve ever seen came from one of my readers. Check out this monstrosity.
H2H mixed points, 12 teams daily moves, non-keeper.
1 point per category won overall each week. W-L-T
R, H, 1B, 2B, 3B, HR, RBI, SB, (TB) Total Bases, A, FPCT, AVG, OPS
W, CG, SHO, SV, ER, K, HLD, ERA, WHIP, K/BB, NH, PG
I just. I hate it. I can see the good behind it — Let’s give players the chance to add value in all sorts of different ways! Let’s put fielding in! Let’s reward big moments! — but it’s all wrong. By putting a stat into a category that you can win each week, you’re putting them on equal footing. So this setting basically says that a stolen base is as valuable as an assist. That a perfect game is the same as a save. First, that’s nonsensical. Second, you’re adding in some stats that are impossible to predict and have very little real-life value. Fielding percentage! Todd Helton was fifth in baseball in fielding percentage last year! Second, putting in things like complete games does strange things to the matchups. Almost every week, CG and NH and PG are ties in this league. You happen to stream Edwin Jackson against the Rays, he walks eight guys against six strikeouts in one of the worst no-hitters in history, and you randomly get two categories rewarded to your team that week for no good reason.
And the last problem with this setting is all the double-counting. I haven’t emphasized that aspect because, to some extent, you’re double-counting all the time. A home run is a hit, a run, an RBI and a home run. In this setting, a home run is a home run, a hit, a run, an RBI, four total bases, and help in average and OPS. Septuple-counting. If you zoom out, maybe the hitting stats aren’t soooo terrible. The easy question is, whom does this system value. And in this case, by counting every hit in the hit, type of hit, total bases, batting average and OPS category, you’re valuing high-contact power hitters most of all. I tell this guy every year to pick up the lowest-strikeout rate guys he can find on offense.
And that’s the point, really. Agree to settings. Then find the players that are best for your settings. It’s a game, and you just have to agree to the rules.
Print This Post