# INDUSTRY FORUM

• chrislamb92

I am writing this post hoping to gather support for the original GLA scoring method draftkings used for CSGO, and not this new horrible one. When I emailed them voicing my concern about the new system I received a response saying they got mostly negative feedback about the GLA system. I imagine that because those of us with brains that function over a 5th grade level who could understand the scoring didn’t feel the need to voice our support for it.

The old scoring system was perfect, it effectively was a player’s:

Score per round * max number of rounds (not including overtime)

This meant that is a team won 16-0, 16-0 in a best of 3 their score would be divided by 32 (to get score per round) then multiplied by 90 (max possible rounds).

Now, DK has decided to give .5 points per round not played within a map, which is much less than what a top player scores per round. Oh, yeah they’re also giving it to both teams. Lose a map 16-0, here’s 7 bonus points for you!

Plus, if you win in 2 maps you get a whopping 5 extra points for not playing the 3rd map! A whole .16 points per round, which is worse than what the worst players on the worst teams average per round.

To illustrate this change lets look at a scenario. Let’s say player A averages 2 fantasy points per round in a match his team wins 16-0, 16-0. Player B averages 1 fantasy point per round in a game his team wins 16-14, 14-16, 16-14. Who had the better game? Obviously the answer is player A.

The GLA format captured this, scoring under the GLA format:

Player A: 64 points / 32 rounds played * 90 max rounds = 180 fantasy points
Player B: 90 fantasy points (no gla)

Let’s take a look under the new format:

Player A: 64 fantasy points + 7 points rounds not played map 1 + 7 points rounds not played map 2 + 5 points (lol) no map 3 = 83 fantasy points
Player B: 90 fantasy points

That’s right, player B now had the “better” fantasy day, all because his team did worse.

Now you may be thinking, but this happens in all sports it’s just a fact of life. However, here is the difference. In traditional sports, the length of the game does not change if players play too well. Coaches may decide to not play their players, but the NBA doesn’t shorten the game because the Bucks are up 30. The mlb doesn’t call a game after 6 innings because the Astros are up by 10 runs.

In CS:GO the game is actually made shorter because your players do well. GLA made sure they weren’t punished for this, the new format falls woefully short.

• jmartin819

I wholeheartedly agree with you chrislamb92. I had been really enjoying how Draftkings was handling Esports game lengths via GLA (CS GO and Rocket League). The math in their new system just simply does not work out. Having to root for a players team to do worse simply because you need them to play more rounds to score higher feels awful and completely takes away from the enjoyment of it. I have also voiced my complaints to DK and hope others who have been enjoying this do the same.

• jhart422

I couldn’t agree more. You’re now rooting for a competitive, maximum map series, which runs counter to dominating the opponent. In the previous format, it would have been nice to breakdown the GLA by map (or in the case of a sweep, the unplayed 3rd map), but at least it was less of a crapshoot than it is now.

• sjobarzanek

no reason to stack anymore on a 3-4 game slate….today pissed me off if there was GLA i was taking down the 10k 6 o clock CSGO so fucking stupid

• chrislamb92

Let’s take a look at how this new scoring impacted some of the scores from yesterday. Before I do that I should note that in my initial post I did not realize that you do get rounds not played for map 3 if you win (or lose) in a sweep, which makes the scoring slightly better, although still terrible.

Yesterday’s slate offered a perfect look into why the new scoring is bad. Cloud9 and Swole Patrol both won 2-0, while nV and Bad News Bears won 2-1, with Bad News Bears needing double OT to win the 3rd map.

Here’s the scoring breakdown for the top player from each of those teams:

Zellsis (Swole Patrol): 62.5 fantasy points over 50 rounds = 1.25 fantasy points per round
Floppy (Cloud9): 61.5 fantasy points over 56 rounds = 1.10 fantasy points per round
Calyx (nV): 83.5 fantasy points over 76 rounds = 1.10 fantasy points per round
dapr (BNB): 77.5 fantasy points over 90 rounds = .86 fantasy points per round

So Zellsis was clearly the top performer of the day, Floppy and Calyx both performed very well, while dapr was a fairly distant third. With GLA their end scores would’ve reflected this. They would have been:

Zellsis: 112.5 (1.25*90)
Floppy: 99 (1.10*90)
Calyx: 99 (1.10*90)
dapr: 77.5 (.86*90)

However, because of the new format the actual scores were:

Zellsis: 87.5
Floppy: 83.5
Calyx: 90.5
Dapr: 83.5 – since maps 1 and 2 went less than 30 rounds he got bonus points, map 3 double OT happened to make the total rounds 90.

Zellsis ends up with less points than Calyx and barely more than Floppy or Dapr!

This doesn’t even get into other players on these teams or players on losing teams. For another example look at the winning lineup from the big \$10. 5/6 players from games that went 3 maps, with the 6th being an outlier game from PwnAlone, who truly lived up to his name while his team got crushed by C9. 0 players in the winning lineup were from Swole Patrol or Cloud9, the two teams who clearly performed the best on the slate.

DK has ruined what was a near perfect DFS product.

• chrislamb92

appears the less than sign is causing my posts to be cut off, editing my initial post (mod if you see this you can delete this comment)

• Otis416

I went from 1200 to 40. Because they gave all of jalepeno players 20 points for 40 rounds not played. They got smashed. But 69 percent had freddy frog. He went 47 and 49. Amd they gave him 20 points. Wtf.

• DraftKings_CM

• ### DraftKings Representative

In CS:GO the game is actually made shorter because your players do well. GLA made sure they weren’t punished for this, the new format falls woefully short.
I wholeheartedly agree with you chrislamb92.
I couldn’t agree more.
no reason to stack anymore on a 3-4 game slate….today pissed me off if there was GLA i was taking down the 10k 6 o clock CSGO so fucking stupid
DK has ruined what was a near perfect DFS product.

Whenever we make a change that our research shows will be positive to grow the game, we hear mixed opinions. This is normal, and expected as we work hard to assure that we understand all the different perspectives from every point a view a group of players may have.

As is the case with many of our recently-introduced DFS options, we have and will continue to test and evaluate for opportunities to make improvements. What this means is that nothing is set in stone until we land at the most optimal solution.

This is standard practice, especially in the early stages of a new DFS product, as we adjust and evaluate different scoring systems until we find the best possible solution. Even then, our work isn’t done, what may seem perfect now could deem to be re-evaluated down the road for myriad reasons (league changes, preventing the game from going stale, etc).

We can’t predict for certain how CS:GO scoring will look weeks or months from now. Perhaps testing calls for a re-implementation of GLA, perhaps we scrap everything and start again from scratch, we don’t know. The only certainty is that we hear your voice, and our game designers are not complacent, they will work until we get this right and then we’ll work to keep it right!

For the latest information on events and when sports are coming back with an updated 2020 sports calendar due to COVID-19 changes, check out our ‘Sport & League Status’ page here: https://www.draftkings.com/about/covid-19/updated-2020-sports-calendar/