Rising to the Fantasy Playoff Challenge

First off, an overdue Happy Holidays and New Year to you all. I spent the last two weeks on vacation in French Polynesia with limited access to football. So instead of being disappointed by Victor Cruz, I was sailing away on a cruise ship; instead of getting beaten by Drew Brees, I was feeling the cool breeze on a beach in Bora Bora; and instead of fretting over whether to start Delanie Walker, I was fretting over my empty glass of Johnnie Walker.

Now back to reality, fantasy games based around the (real) NFL playoffs are a peculiar animal – not nearly as exciting as the fantasy playoffs but a little more rewarding than say, playing in the consolation ladder of said fantasy playoffs.  But there are only a few weeks left of football, so this is your last chance to feel smarter than your leaguemates. Well, unless you play in a fantasy Pro Bowl league, in which case you are a greater man than I.

NFL.com’s Fantasy Playoff Challenge boasts itself as the #1 playoff fantasy game, and considering the amount of promotion they are giving it, it will likely remain in top spot for some time. So I'll focus on this particular game and format for the purposes of this article

You can learn more about the rules on their site, but here is a brief overview:

• The scoring system follows a typical non-PPR format
• Your roster consists of only 1 QB, 2 RB, 2 WR, 1 TE, 1 K and 1 DEF.  There are no bench spots
• You create the best roster you can before the start of each round and your points accumulate each round (so no head-to-head).  You can change your entire roster each week if you want, but as you’ll see below, it may not be wise to do so
• There is a multiplier effect that awards you more points the longer you have an active player on your roster.  So if you pick Colin Kaepernick before this weekend and he takes the 49ers to the Superbowl, you’ll receive 1x his points this week, 2x next week, 3x the following and 4x in the Superbowl.  Likewise, if you’ve neglected to pick up Julian Edelman until the AFC Championship and he goes on to the Super Bowl, you’ll get 1x his points in the AFC title game and 2x in the Super Bowl.  Clearly, it pays to pick players that can make multi-round runs.
• Finally, for the teams on bye – if you select any of their players before this weekend’s games commence, they will begin next week on your roster at a 2x multiplier.

That last point is a critical one and a major aspect of the research below.  If you choose a player on a bye this week– Tom Brady, for example – you won’t get any points from him this week, but double the points next week.  Alternatively, if you pick a QB playing this week – say, Andy Dalton – and he advances, you’ll get 1x his points this week and 2x his points next week (assuming you choose to keep him on your roster).  If you dropped him for Brady, you’d have Brady at only a 1x multiplier for next week.

Now, the basic strategy is to pick which teams you think will advance to the Super Bowl (or at least the Conference Championships) and load up on their players.  Most pundits and sports books favor a Seattle-Denver Super Bowl, so you could simply fill up on them and be done with it.  But things never go according to form – the last 2 Superbowl winners (BAL. NYG) were #4 seeds and the year before (GB) a #6 seed.

I’m not going to pretend to know which lower seed – if any – will make a magical run over the coming weeks.  But what I aim to do is put together a model that will show you what ought to happen, based on season averages and current probabilities of success.  We’ll then revisit the model each week with a fresh set of probabilities and determine whether to keep or swap.

Remember that this type of fantasy game has a fairly limited number of outcomes.  There are a limited number of players and it’s highly likely that you’ll have many of the same players as the other people in your pool.  So every edge counts.

Note, for each player, I’m using their average fantasy ppg for this season.  It’s dangerous to rely on too small of a sample size – like the previous 5 weeks – or to put too much stock in momentum.  It's too large of an analysis for the purposes of this article, but if you're interested in why, I recommend Scorecasting: The Hidden Influences Behind How Sports Are Played and Games Are Won by Tobias Moskowitz and L. Jon Wertheim.

I am also using Football Outsiders’ probability of appearing in the Conference Championships that are based on their current weighted DVOA ratings.  This percentage gives me a robust and unbiased probability to weight against their expected point output.  We can then look at both their total expected points as well as the probability of attaining those points to see what options we have.

Finally, I’m not showing you every possible QB, RB, WR or TE.  I’ve only taken the top performers on the season and compared them against the player(s) with the highest likelihood of advancing.  If there is a player not detailed below, you can be sure that they have a lower fantasy ppg average and/or a lower probability of making their Conference Championships.

Alright, enough talk, let’s see it in action.

Quarterback

In this scenario, we’re picking one QB and sticking with him until the Conference Championships.  What I wanted to see was whether it was better to ‘wait’ for Peyton Manning or go for someone like Drew Brees or Andy Dalton who are playing this weekend.

As you can see, if you knew Brees would make it to the NFC Championship, you’d pip the other owners in your league, who will likely all be rostering Manning.  But the probability of that happening is very low – consider that the Saints are 3-5 on the road (where they score about half as many points as at home), travel to a cold Philadelphia this week and then would face Seattle should they prevail.  The separation from Manning is only 9 points anyway and that’s if both players played exactly according to their average.  Considering that the Saints likely have more difficult matchups to get to their Conference Championships than Denver, it’s likely Manning would make up that difference and potentially even surpass it.

Note that Russell Wilson and Nick Foles are also listed for comparison sake.  Even if you’re certain Seattle will at least make the NFC Championship, Wilson doesn’t produce enough points to really separate from the pack.  If you really don’t believe in Manning, you’d be better off rolling the dice with a Foles or Brees type and then switching off of him if/when he gets eliminated.

Now, what if we went with Brees this week and then switched to Manning; or went Brees for 2 weeks before switching to Manning?

In both scenarios, you’d be better off just sticking with Manning from the start.

Running Back

This table is a little more interesting.  Clearly, Jamaal Charles and LeSean McCoy represent tremendous point potential over steadier options like Marshawn Lynch and Knowshon Moreno.  Even if they only get to the second round, they should still be ahead at that stage.  The problem, of course, is whether they can advance beyond this week, let alone next.  The Chiefs are struggling and a slight underdog to Indy this weekend.  Philly are favored over New Orleans but would play Carolina next week, who are one of the top run defenses in the league.  In fact, if Philly faced the Saints, Panthers and Seahawks in their run through the playoffs, they would be facing the 8th, 5th and 2nd stingiest defenses to opposing running backs in fantasy this season.

Still, that point differential is enticing.  What if we adjusted the probability percentage of reaching the Conference Championships?  Or to put it another way, what percentage of chance would Charles and McCoy need to match Lynch and Moreno’s weighted value?

For Charles, you would need to double his percentage (to 41%); while for McCoy, you would need to increase it to 49%.  Would you say there’s a 50/50 chance McCoy makes it to the NFC Championship game?  I sure wouldn’t, but that’s what you need just to match the weighted values of Lynch and Moreno.  For Charles, it’s ‘only 41%’ but considering they aren’t even a favorite in the wild card round, that’s a pretty tall order.

What if we use the hybrid approach of starting Charles or McCoy this week and then making swap to Lynch or Moreno in Week 2 or 3?

Here, the only option that potentially makes sense is to start Charles this week and next and then swap to Lynch (or Moreno) – although it’s a slim advantage over simply going with Lynch or Moreno from the start.  Remember, you only need the player to play that week, not advance.  So in this scenario, you’re only asking KC to beat Indy in order for Charles to register 2 weeks of fantasy points.  If they lose the following week to NE or DEN, you would then swap in Lynch or Moreno for the Conference Championship week and still be on pace to come out ahead (albeit only slightly).  But the added benefit is that you get the possibility of Charles winning next week – as remote as you think it may be – and then would be faced with an opportunity to take him in the AFC Championship with a 3x multiplier.

Also note that Indy are roughly league average (13th best) in terms of fantasy points conceded to RBs, while Denver and New England are 18th and 20th respectively.  So there will be opportunity to run.  But this strategy all comes down to whether you think KC can advance, because a one-and-done puts you about 9 points off of simply rolling with Lynch or Moreno.  That might be too steep a price to pay for a fling with Charles.

Wide Receiver

This one is pretty cut and dry.  Your best bet is to roll with Demaryius Thomas and Eric Decker until Denver loses.  Unless you believe they are in for an early exit and you can predict a long run by Cinci or Green Bay (or a long run and star performance from someone like TY Hilton or Keenan Allen), the odds are just not in your favor.  You’re better off taking Thomas and Decker and if they lose, simply jumping on the surprise team from that point forward.

Tight End

This result had my interest piqued as I was punching in the numbers.  Was Jimmy Graham that dominant that even 1-2 games with him would be worth it over some of the others?  Well, it would appear not.  Even Vernon Davis comes out slightly ahead in weighted value. In fact, if you believe that the 49ers will go far this year (remember, they went to NFC Championship two years ago and the Super Bowl last year), you would get a nice boost by choosing Davis over Julius Thomas due to Davis playing one extra game.

Now what about the hybrid approach?

There is a bit of Charles-esque scenario at work here.  If you believe New Orleans will prevail in Philadelphia and you started Graham both this week and next, you’d be on pace to roughly match the output that you’d get from simply rolling with Thomas from the start.  But you’d also get the upside of a potential second victory by the Saints that would give you the opportunity to play Graham with a 3x multiplier (and maybe 4x in the Super Bowl).  Couple that with the fact that Thomas missed games to injury 4 weeks ago and has been a little inconsistent since his return, and there’s a window of opportunity for Graham fantasy believers.

Personally, I like this tactic as I’m worried that everybody is a little too high on Philly right now and perhaps underestimating Brees and Sean Payton.  I’m not convinced that Thomas will be a top TE during this playoff run either.  The playoffs never go according to form so if I'm going to pick one player from an underdog, I figure I might as well pick someone who is head and shoulders above his peers.

Of course, you'll have to make similar decisions for your team. The analysis above strictly follows the weighted DVOA rankings but recent history suggests that the top ranked teams aren't assured long playoff runs. So if you can pick the darkhorse that will go on a long run, you just might have that single advantage that takes you to the top in your league.

We'll revisit this next week, apply some updated numbers to it and determine which players from the past weekend you should keep for the multiplier effect and which you should drop.
[/am4show]

–>