Going into the inaugural season of Daily Fantasy CFL contests on DraftKings, no one really knew what to expect from the game-play. The scoring was similar to NFL and CFB contests, but rosters were obviously different and the slates were all four games at most spread over three to four days. It was in many ways vastly different from other football offerings. The CFL contests started quite large, but after about six weeks really started to dwindle in size as NFL training camps opened up and Seth Yates preseason NFL projections became the hottest ticket on the Twitter machine. Let’s be frank: The CFL will never be the NFL. That’s okay. CFL contests don’t need to be able to support a Milly Maker to be a lot of fun and provide some good winnings. The closest comparison we have elsewhere may be NFL Wildcard playoff slates in which there are a similar number of games and teams involved. The biggest stars will be on almost every roster, so it will come down to getting the supporting role players right to come out ahead. You have to be comfortable with having entry fees tied up for half a week at a time. There’s an awful lot of variance involved in small slate contests like this, so the strategies involved in coming out on top have to be different than your standard NFL slates.
The biggest change in CFL contests on DraftKings this season is in roster construction. The 2016 had rosters that consisted of 1 QB, 1 RB, 3 WR, 1 RB/WR Flex, and 1 D/ST unit. This season they have dropped one WR from the lineup and replaced it with an extra Flex position. The added flexibility may not really help in creating optimal lineups, but hopefully will allow others to put three RBs in their lineups. I’ll look at lineup construction a little later on though.
CFL contests on DraftKings started out fairly large and had a lot of people excited about the game, but as the NFL got rolling, CFL interest started dwindling. There was a lot of good money to be made in early season cash games as there were a lot of players with very little CFL knowledge involved, but by mid-season there was little edge to be had as the gash games were primarily filled with those who knew what they were doing. It appears that the week one contests will again be quite large, but are essentially all 20 max entry GPPs. Mass Multi-Entry strategies won’t likely yield the same returns as you might find in other sports as it will be much more difficult to control exposures effectively with fewer entries. So what’s a person to do?
As I wrote earlier, the small player pool will lead to massive overlaps in rosters, very high ownership of dependable players, and small margins between those who take down big GPPs and those who don’t even cash. With small slates like this I would suggest limiting your entries in cash games to significantly less than what you may play in more mainstream sports like NFL, MLB, or NBA. If you are going to play cash games though, early in the season will be the time to do so, hoping to attack weaker players who are new to the game while there is still some edge to be had. Often the difference between winning a H2H will come down to having the right flex players as there will be quite a bit of consensus on who the top QB, RB and WR of the week will be. By the end of week six, I will probably be out of cash games altogether and sticking with GPPs. In GPPs the winning strategy will likely be taking the contrarian stance by avoiding too much exposure to popular plays and hoping that they flounder. In small slate GPPs variance has a much larger effect on the entry pool, so keeping your exposures in mind will be very important.
The CFL is a drastically different league than the NFL. The rules, the size of the football, and the dimensions of the field may be the least of the differences though. While the NFL has franchise players and rewards teams that build through the draft, the CFL has rosters in constant flux with few contracts being signed for more than 1-2 years. Star players are born from open tryouts in the off-season, and major contributors arise throughout the season as opportunities due to injury or ineptitude come around. There are few established statistical providers, a fantasy analyst community in its infancy, and no vast deluge of weekly projections to work from. Chris Kay and I will be putting out two articles a week: one with some strategy or statistical analysis and one with weekly rankings for DraftKings contests with cash and GPP plays. We are working on our own projections model for CFL statistics, but it’s still in Beta at this point. Our articles will certainly draw on that data, but they aren’t available for public consumption yet. Playing CFL contests and building lineups is quite a bit different from many other leagues as very few competitors in the weekly contests will simply be downloading subscription based projections, plugging them into an optimizer and just rolling with their computerized lineups. A lot more hands on work is needed to succeed with CFL DFS.
Unlike the NFL and college football, the CFL is a passing based league. This is driven by having only three downs to gain ten yards, a field that is ten yards longer from goal line to goal line, 12 yards wider, and featuring end zones that are 20 yards deep. This means that even in the redzone there is ample space to run routes and get open, unlike in the American game where everything is compressed in the scoring area. The big field makes a big difference. In general, scoring is higher, passing yards and touchdowns are more common, consistent production from running backs is rare, and defences are hard pressed to make stops even before coaches throw replay challenge flags begging for pass interference penalties (which they often get).
As I mentioned before, roster turnover, not just season to season, but week to week is the norm. Players regularly lose starting jobs and roster spots to injury. At any time in the league there are usually two to three dependable quarterbacks, a couple of every down running backs, and a plethora of receivers to spread the ball around to. Don’t try to get to know the names too well: they’ll mostly be different by mid-season. Success will likely come through keeping up to date on depth charts as they come out twenty-four hours before kickoff. Using the late-swap feature effectively may be the difference between winning and losing weekly in slates that span three to four days.
The personnel involved in the game-play are quite a bit different from American football as well. A standard formation will include the quarterback, one running back, and five receivers on offence. There has also been a move to bringing in specialized mobile QB’s in redzone packages to take advantage of the running lanes in close. This has been limiting the upside of some top QB’s in Edmonton, Calgary, and Hamilton. Many seasonal leagues have actually switched to “Team QB” to limit this effect.
Coming into any season, there are generally two or three teams with well-established quarterbacks, two or three rolling with an aging veteran or promising but unproven talent, and two or three that have no idea who the quarterback will be from week to week. Coming into this season you can likely count on Mike Reilly, Bo Levi Mitchell, and Jonathan Jennings on a weekly basis. Zach Collaros, Trevor Harris, and Matt Nichols are a slightly less dependable second tier. This then leaves the mess of what’s left of Darian Durant, Ricky Ray and Kevin Glenn in the bargain bin to pick away at. Many mobile quarterbacks will come and go throughout the season using the large field to run and create offense. Very few of these players will last more than a couple weeks as starters. Don’t get caught up in the hype.
It’s not at all uncommon to just field a six pack of receivers with an empty backfield in the CFL. Given the number of receivers involved in a game, it is often difficult to predict market share of targets beyond the top receiver for a team. The distribution is often spread throughout six or seven receivers. For some teams like BC and Ottawa, it will be hard to determine who the top target will even be though as they have so much depth at the position. Look at Adarius Bowman, Luke Tasker, Naaman Roosevelt, Duron Carter, Darvin Adams, Kamar Jorden, and DaVaris Daniels as dependable weekly plays who should get consistent targets. One of the trickiest things to decipher early on this season will be the target distribution in BC and Ottawa where they have a ton of high end talent, and offences that will create a lot of points but also have a lot of questions surrounding who will actually get the ball. Chris Williams, Emmanuel Arceneaux, and Bryan Burnham in BC can’t all be WR1’s weekly although they will likely be priced as such. Similarly, Greg Ellingson, Brad Sinopoli, Kenny Shaw, Diontae Spencer, and Tori Gurley have all shown flashes of WR1 talent, but can’t possibly all produce at that level in the same offence. Deciphering the right play on a given week will be difficult. The difference in winning a GPP each week will likely come down to who can pick the right secondary targets on the highest scoring team of the week. Focusing on rostering players with a high number of targets, and looking for high efficiency rather will help create winning lineups.
Running back is an interesting position in the CFL as most top performers only stick around for a season or two before going to be a third stringer in the NFL. A practice roster minimum salary in the NFL will still be four times the top money paid to a starting RB in the CFL. With the game being so much more passing based, pass protection skills are often perceived as more important than actual rushing skills in potential running backs. Workhorse running backs in the CFL don’t exist in the same way as the NFL game. No one gets 20+ touches a week. There are a few backs, such as Andrew Harris and Jerome Messam that will get consistent touches but they come at a premium price. Both are aging and will certainly miss time to injury. Jeremiah Johnson and Brandon Whitaker make up a slightly less consistent second tier. After those two, it will be somewhat of a crap-shoot from week to week trying to determine touches both in the passing and running games for every other RB in the league. You can expect to see numerous different starters in Saskatchewan, Edmonton, Ottawa, Montreal, and Hamilton. It’s also good to note that Saskatchewan, Hamiton, and Montreal ran the ball less than 24% of the time last season, so their RBs rarely were good values even at very cheap price tags.
All this being said, going cheaper at RB certainly seemed to be the optimal way to build cash game lineups through the 2016 season. There was some early season buzz around the value of kick returners as RB plays after Chris Rainey had a couple big special teams games early in the season, but that theory was soon discounted. The players with game winning upside were usually at the WR position, and in order to get multiple elite WRs in your lineup going cheaper at RB – ideally in the $4-6k range – led to the best outcomes. When value plays emerged at QB as well (ie. Matt Nichols) it became possible to even play three elite WRs. In regards to defence, the Winnipeg defence had an unbelievable mid-season run putting up astronomical scores through multiple turnovers and TDs, but it’s unlikely we’ll see a defence like that again. As Chris found through his analysis, going cheap at D/ST is usually the optimal play.
Hopefully this gives you a good intro to the Canadian game. It’s a lot of fun to watch and play along with. Come join the CFL conversation in the Forum, find Chris and I on Twitter at @RealestChrisKay and @benyamen, and buckle up for the start of another wild season in the CFL.