# Devising A Strategy To Win The DraftKings National Championship | Part 2

In Part One, we outlined the structure and rules for the DraftKings Sports Betting National Championship being held November 5-7, 2021. We also helped calculate your Expected Value (EV) for the tournament. More importantly, devising a strategy to win the DraftKings National Championship is a key component to realizing any +EV in your approach to the tournament.

In Part Two, we take a look back at the 2019 tournament to see what we can learn from player behavior. Then we devise some strategy approaches and simulate their effect over 10,000 tournaments to get a sense of how parts of the tournament will play out. Lastly, we’ll come up with some key takeaways that the Unabated reader can use to find their path in the tournament.

## Learning From The 2019 DraftKings National Championship

Going into the event in 2019, some thought the winning bankroll would be in excess of $500,000. That requires a 100x return, but with 259 people, if everyone bet a parlay that paid 100/1, we’d expect at least one person to hit, even if true odds were 150/1 or higher. The fact that the winning bankroll barely broke $100,000 says a lot about how poorly most entrants played it. Part of that could be explained by the human element of risking a real money bankroll.

One of the changes for 2021 makes the payouts extremely top-heavy. 50% of the prize pool goes to first place. Clearly, the value proposition is to play for first. If you’re not comfortable losing everything 90% of the time, then you should not be playing this event. Full stop. You will be dead money.

Contestants in the 2019 SBNC made two big mistakes:

- Being too conservative
- Being too diversified

These are great qualities in betting and investing, but they are cardinal sins in the DKSBNC.

## Devising Your Tournament Strategy

Early tournament strategy is the most clear-cut thing and easiest to quantify. You have one objective and one objective only at this point. You are playing for first place. Note that in extreme cases such as 90% of people going all-in on longshot parlays, this is not true, but unless everyone competing in the DraftKings National Championship reads this article, this situation is completely unrealistic. There are two good options for early tournament strategy:

- Go all-in on one bet.
- Do nothing.

Going all-in on Friday night was a popular strategy in with DFS players in 2019. 42 of 259 entrants went through their entire bankroll without winning a bet. 12 did so on one bet. But only three of those 12 were +160 or higher and none was higher than +200. Most were actually laying a small price. Still, this is a solid strategy, especially if it’s not employed by that many people.

Another strategy that seemed very popular in 2019 was the **diversified longshot all-in**. Divide your bankroll into many different longshots. The aggression is good, but the diversification is fundamentally misguided. Variance is your friend. You are rooting for a tail outcome. An extreme outcome. The larger your sample, the more likely results end up closer to the fat part of the bell curve. Minimizing downside should not be a concern.

One entry placed 27 bets at an average price of 16/1. They unfortunately lost them all, but each was on average risking $185 to win $2,960. The most likely outcome for them was winning just one of them. The bettor was likely employing a round robin strategy. If they were uncorrelated, the chances of massively increasing their bankroll was very slim. They sacrificed upside to limit downside (and despite that busted, but that’s not the point).

Put simply: this is a bad strategy. Granted, it’s better than some others, but it is not in any way optimal.

## Round Robins? Good or bad?

Many players will be tempted to round robin their parlay bets. If they have 7 teams they like, they’ll do a round robin of 5-leg parlays where you bet all 21 permutations of 5-leg parlays across those 7 teams. What’s the advantage of a round robin? It reduces your variance. If all seven legs hit you’re swimming in it. But you’d have made the same amount if you bet it all on one 5-leg parlay, which is four times more likely to hit than the 7-leg parlay.

In this round-robin example, assuming someone bets their entire $5,000 on this round-robin with -110 odds on each leg and 50% probabilities for each bet, this is what the payout schedule looks like for the round robin:

Result | Probability | Bets Won | Bets Lost | $ Won | $ Lost | P/L |
---|---|---|---|---|---|---|

5 of 7 | 16.4% | 1 | 20 | $5,799.76 | $(4,761.90) | $1,037.86 |

6 of 7 | 5.5% | 6 | 15 | $34,798.57 | $(3,571.43) | $31,227.14 |

7 of 7 | 0.8% | 21 | 0 | $121,795.00 | $0 | $121,795 |

So in this case there’s a 6.25% chance of increasing your bankroll by a bunch. But you could also just place a single 4-leg parlay with your $5,000 that would give you a 6.25% chance of winning $61,416.57.

Remember that variance is your friend. While it may feel safer, *safe is not what you want!*

There is a second component to this. More bets means more of a house edge. Even if you find +EV bets, there is a tradeoff between volume of bets and expected return. If I placed 3 bets per week on college football, I’d only bet my three biggest edges and would expect to fare much better than if I made 30 bets.

## Simulating The DraftKings National Championship

As a thought exercise, we created a basic simulation of the DKSBNC. Some simplifications were made. We divided the tournament into four distinct betting sessions (Friday night, Saturday early, Saturday late, and Sunday). In each session, each entry bets based on a predetermined strategy. All bets (and legs of parlays) have 50% win probability and -110 odds offered. We assumed 400 entrants and divided them into 5 groups of 80:

**DOUBLE UP**: The entrants will bet their entire bankroll on a straight bet in each of the four sessions. Their goal is to double up their bankroll 4 times over.**DIVERSIFIED PARLAYS**: In each session, these entrants bet their entire bankroll on 10 4-leg parlays. They believe parlays are the path to victory but they want to spread their money across many different parlays.**CONSERVATIVE DIVERSIFIED PARLAYS**: Same as diversified parlays except if the entrant has doubled up, they only wager 80% of their bankroll during the next season. This takes into account the human element when wagering real money. Keep a little behind to be safe.**DIVERSIFIED STRAIGHT**: These entrants don’t trust parlays so in each session, they bet their entire bankroll on 10 straight bets at -110 odds.**WAIT TO FIRE LATE**: These entrants do nothing for the first three betting sessions so they can wait and see what it will take to win the tournament on the last day. In the last betting session, they go all-in on a parlay of either 4, 5, 6, 7, or 8 legs (determined at random).

We ran 10,000 simulations under these conditions, and the clear winner was the **WAIT TO FIRE LATE** strategy, where all but the 8-leg parlay showed positive returns on average. Here is the average EV with prize money factored in for each approach in our simulation:

- DOUBLE UP: +$159
- DIVERSIFIED PARLAYS: -$3,657
- CONSERVATIVE DIVERSIFIED PARLAYS: -$3,770
- DIVERSIFIED STRAIGHT: -$4,428
**WAIT TO FIRE LATE: +$3,480**

Clearly, fewer bets is preferable to more bets if the house has an edge. That’s nothing new.

## What If Entrants Gravitated To Better Strategies?

What would happen if most entrants played the fewer bets strategies (1) and (5)? We re-ran the simulations with 35% of entrants in each of (1) and (5), and 10% in each of the remaining three. EV for each strategy was as follows:

- DOUBLE UP: -$1,990
- DIVERSIFIED PARLAYS: -$4,518
- CONSERVATIVE DIVERSIFIED PARLAYS: -$4,319
- DIVERSIFIED STRAIGHT: -$3,313
**WAIT TO FIRE LATE: +$1,571**

Interestingly, the returns on every strategy except the Diversified Straight bet strategy went down. There is an inflection point where the Double Up strategy becomes a poor strategy. And if we took this farther, we’d find that there’s a point where a very conservative strategy (even a do-nothing strategy) becomes profitable.

Every strategy becomes worse if it’s more widespread. Being contrarian is good. If you notice that a large percentage of people are going all-in early, there’s value to waiting. If everyone is conservative, there’s value to being aggressive. Where there isn’t value is the middle ground. You should be playing for first place or for the minimum cash. In many respects, the optimal strategy for use in the DraftKings National Championship won’t be known until the tournament is significantly underway.

Amazingly, you’ll have the opportunity to know how the tournament is progressing before you even decide to enter. Registration will be open until 6pm ET on Saturday, November 6th. A full 24 hours after the tournament starts. You can scope out what early tournament strategies are being used and how you can exploit that. This brings us to late tournament strategy.

## Late Tournament Strategy

This is much harder to simulate, as behavior really will depend on a number of factors. Because your bankroll is active, people at the top of the leaderboard entering Sunday will likely be more conservative than they should be.

If you grow your bankroll to $80,000 entering Sunday, it’s mentally difficult for most people, especially if they’re not high-stakes bettors, to accept that they’ll likely lose it all if they play optimally. Having the option to play the other side of a bet you make in the contest can help you lock in profits, but not everyone in this contest will have $80,000 sitting around to bet. So, to reiterate, people will be more conservative than they should be.

## Learning From 2019

In the 2019 contest, Rufus went all-in during the 1pm Sunday game on a straight bet (Patriots -3) in an attempt to double-up his bankroll. He was sitting in 5th place at the time, trailing the leader by less than $10,000. None of the leaders were aggressive in that early time slot, and had his bet been graded in time, he would have been in position to go all-in on the late game at a price of -175.

If you’re at or near the top of the leaderboard entering Sunday, you are going to have an interesting calculation to make. What do you think the winning bankroll will be? If people are more conservative (as we’ve theorized), it’s not going to be as much as you think. There’s a risk/reward calculation here. You don’t want to play a bigger longshot bet than you have to, but you also don’t want to end up underestimating the winning number and not giving yourself a chance. There’s no easy answer here. But this is also the fun part!

## The Expected Value Framework

There is a framework to use. It is our old friend, Expected Value. We have to make a lot of assumptions, but you can evaluate a strategy this way.

We need to know:

- Probability that our bet(s) win and what our bankroll is in that case
- Conditional on our bets winning, what’s our expected prize money? Break this down by probability of winning and probability of not winning (and expected prize money if we don’t win)
- Bankroll and prize money conditional on our bet(s) losing.

In the 2019 DKSBNC, Rufus made some back-of-the-envelope calculations of his EV, starting Sunday in a fourth-place position. For each strategy he thought about the probability of winning the contest if his bets won. That required assumptions on how aggressive other contestants would be, both those ahead and behind him. The numbers were just estimates, but it gave him a framework to make a “professional guess.” In the end, he actually severely overestimated how aggressive other participants would be.

## What if you’re out of it?

Just like there are two acceptable early-tournament strategies, we believe there are two acceptable late-tournament strategies.

- Play to win
- Sit pat and/or go for the minimum cash

The way the prize pool is structured, there is a steep drop-off from first to second place, but not a big drop-off after that. If you’re in 18th place entering Sunday, you should not be playing to get to 10th. You should either go all-in to try to get 1st place, or try to guarantee the minimum cash.

## Trying To Move Up A Few Places Is Simply Not Worth The Risk

If you’re out of the money, you still have these same two choices. You can still play to win even if you sat out both the Friday and Saturday sessions. The fact that there are so many NFL and NBA games to choose from means it’s less likely there’ll be that much overlap if you make a YOLO parlay. In the 2019 contest, DraftKings accepted parlays to win north of $200,000. If you can bet $5,000 on a parlay at 100/1 odds, you are still very much alive. In fact, you are in a decent position given that you have a much better idea of where the winning bankroll will be than you did no Friday or Saturday.

Now if one of these all-in entries hit a big parlay and the lead is $400,000 going into Sunday, it may not make sense to go for it. You may have a 1-in-125 chance of hitting your 100/1 parlay but if you would only have a 25% chance of winning if it hit, that’s not good value.

Probability of bet winning (p)= *0.8%*

Bankroll if bet wins = *$505,000*

Probability of first place if bet wins =* 25%*

Expected prize $ if bet wins and results in first place = *$1,000,000*

Expected prize $ if bet wins and does not result in first place = *$100,000*

EV = *(0.8%)*($505,000+(.25*$1,000,000+.75*$100,000))-(99.2%)*0 *= **$6,640**

## Min Cash Strategy

This is better than standing pat at $5,000, but there is another option available. The **go for the minimum cash** option. Last SBNC, the min cash was approximately $8,900, but I’d expect it to be higher this year, as people (should have) learned from the previous event. It’s important to have an idea of what that minimum cash number is and a best guess of what the probability you will cash if you hit that number.

Let’s evaluate the EV of going all-in on a +125 underdog. The minimum cash is $10,000 for 21st-25th place. $15,000 goes to 16th-20th.

Probability of bet winning (p) =*43%*

Bankroll if bet wins: *$11,250*

Probability of cash if bet wins = *75%*

Expected prize $ if bet wins and results in a leaderboard prize = *$12,000 (weighted between $10,000 and $15,000)*

EV = *(43%)*($11,250+.75*$12,000)-(57%)*0* = **$8,707**

Based on these assumptions, at this point the best move is to play for the minimum cash and not for first place, but both are better than just holding your $5,000.

Based on this logic, you can actually work your way backwards to determine what conditions make the **wait and see** strategy a good one. Clearly, it’s not good in this case, but if, due to the aggressiveness of others, the minimum cash ended up being $7,000, the wait and see strategy (going for the min cash) is a profitable one. In that case you could lay a -200 favorite on Sunday and even if the true probability was only 65%, your EV would be = *65%*(7500+10000)-35%*0 *= **$11,375**, so your expected profit on my entry would be **$1,375**.

Also remember that the wait and see strategy isn’t a binary thing. You gain more information the longer the contest progresses, but you sacrifice time to bet. This brings us to the question we get frequently. Will you be entering the DraftKings National Championship? Unfortunately, Jack won’t be participating since DraftKings closed his account a while back. Rufus’ personal plan is to wait until Saturday afternoon to assess whether he plans on entering. It’s clear Rufus will be using his own Wait and See Approach.

## Projecting The Winning Bankroll

The other question we received most often is “What will it take to win the event?” The two things that matter the most in determining strategy are projected winning bankroll and projected bankroll for 25th place (minimum cash). Projecting those numbers is difficult. There are some thought exercises you can do, but our simulation gives us a lot of insight as well. We expect there to be an inverse correlation between the two (controlling for number of entries).

In our simulations we used 400 entries and the average winning bankroll was **$283,778**, but it varied widely (standard deviation of **$159,769**). Our average min cash bankroll was only **$7,176**. We believe that is unreasonably low, since our simulation is modeling a level of aggression you won’t see, especially among entries at the top of the leaderboard, which is leading to a higher number of entries busting out than we will probably actually see.

It’s interesting to note that 25th place in the 2019 contest had over $16,000 entering Sunday, yet fell to under $9,000 at the end of the contest. Part of that may have been due to a very tight leaderboard, with the leader having a bankroll far less than one would expect. This incentivized more people to go for broke (the right strategy in my opinion).

If the 2021 tournament has fewer entries than our simulated 400, you are likely to see a lower winning bankroll number and a lower minimum cash threshold. More entries will lead to a higher max and higher min cash threshold.

## Key Takeaways

- Read the House Rules.

Even though we tried to outline the rules and structure of this tournament. Take a few minutes and read the full rules of the tournament so you understand exactly how it will be played. After all, with $10,000 at risk, you better make sure you understand everything.

- Variance is your friend. Diversification is your enemy.

When you’re deciding which bets to make in this tournament, remember that key phrase. You should always be looking to make fewer bets not more bets.

- This is not a handicapping contest.

In this contest, strategy matters more than handicapping skill. Don’t be afraid of -EV wagers if those wagers will present an opportunity to catapult to the lead late in the tournament. The winner may be crowned the 2021 Sports Betting National Champion, but in the end they were really just the King of Variance one weekend in November.

- The middle ground is a graveyard.

Either be all-in early or wait, but don’t try to take a middle ground. Spreading out your action over the course of the weekend is very much sub-optimal.

- The optimal strategy depends on everyone else’s strategy.

Every strategy is worse the more people are using it. Identify if any strategy is being played predominantly and be prepared to be contrarian.

- Play for first place. That’s where the equity is. If first place is out of reach, play for the minimum cash amount.

There will be many people who will take comfort in aiming for 2nd through 15th. Don’t be one of them. With 50% of the prize pool wrapped up in first place, you need to be aiming for that. Only when that is not a possibility you can consider shooting for a min cash.

- Use an expected value framework to make strategic decisions late in the tournament.

Give yourself time to evaluate your bet decisions in the last betting session on Sunday. Frame them out by EV. Being pressed for time will force poor decision-making.

If you choose to enter, and remember you can late register up until Saturday 6pm, we wish you good luck – or should we say “Positive Variance!”