Creating the Overwatch Fantasy League
What if I told you that you could be the leader of an Overwatch League team? What if told you it only cost you $20, instead of $20 million? What if I told you you could draft Surefour, TviQ, Kaiser, Ryujehong, and Mek0 all to the same team? Enter: Fantasy Overwatch.
Fantasy for Dummies
Fantasy leagues have a storied history dating back to the 1960s and they are a great way to provide supplemental entertainment and improve actual engagement to the sports and esports they emulate. The idea is simple: league of fantasy owners (usually group of friends) come together to draft individual players from “real” teams onto “fantasy lineups” based on how each owner expects these individuals to perform throughout the season. Owners might draft Surefour, Seagull, and Birdring onto their team even though they play for entirely different teams in real life. In this way, fantasy leagues can drive individual story lines and create strong player-fan connections. Many fantasy leagues add a dash of gambling too – a $10-$20 buy-in for all owners makes the stakes all the more enticing.
Above, you see how a typical fantasy matchup might look in Overwatch based on the system that I will lay out below. Good drafting and knowledge of the system you play in determines success in fantasy, so let’s begin theory-crafting what an OWL Fantasy League might look like.
Defining the rules for creating Fantasy Statistics
There are a couple boxes to tick when developing the statistics and ratios to create a new fantasy system. The stats tracked must be compelling, they must be easily understood, and they must be mostly consistent. But, they need not actually directly correlate to the individual player’s winrate. Indeed, a valid strategy is drafting the best player on the worst team – banking on the assurance that that player carrying his/her team and causing a boost to their individual stats in response.
Blizzard will ultimately be the creators of the OWL fantasy system (in fact, they already have plans in motion ), because they have access to all of the raw data that will be needed to feed into actual fantasy points. For this exercise, I will use estimated numbers – essentially napkin math calculations – to define the framework through which Blizzard might base their own system. I might not have all of the stats, but I think I can create a pretty comprehensive system regardless. As I explain each fantasy statistic, I will be making comparisons to American Football’s fantasy leagues which I believe can serve as a great starting point for the OWL fantasy system.
First, let’s look at yardage:
Yardage vs. “Throughput”
One of the simplest statistics from football to understand is yardage, or yards. For you EU/non-American folks, a yard is roughly a meter and it is the unit of measure by which progress is determined in American football. For Overwatch’s “yardage” fantasy statistic, I propose combining Damage, Healing, and Damage Blocked/Absorbed into one combined total. All of these stats accumulate in a similar way as yardage: the longer a game goes on and the better the player, the higher they climb. Every hero in Overwatch also has the ability to accumulate one or two of these three stats, which means it will be a very consistent source of fantasy points. For simplicity, I’ll refer to this statistic as “Throughput”.
“But Captain,” you might be wondering, “Throughput will be very different total numbers depending on the hero! You can’t expect a Tracer to deal as much damage as a Soldier 76!”
You would be right! But this is a problem that fantasy systems have already solved by adding scaling factors to the raw yardage statistic. Quarterbacks, since they accumulate yardage faster, gain fantasy points at a rate of one point per 25 yards compared to running backs who gain a point per 10 yards – a pretty big difference! Now, let’s apply this to Overwatch to create a situation where a high-performing Soldier 76 might end up with a roughly similar score as a Reinhardt.
I am going to use Overbuff’s hero page to simulate some numbers that might roughly align with OWL-level players.
Starting with Soldier 76, I found that 90th percentile Soldiers average ~18500 damage per game. We also have to take Soldier 76’s healing into account fill out his raw statistic total: 89th percentile Soldiers average ~2500 healing per game, giving us a total of 21000 Throughput per game for a high-performing Soldier 76. Now, let’s look at Reinhardt.
90th percentile Reinhardts deal ~11000 damage on average per game, and 89th percentile Reinhardts block ~23,000 damage on average per game for a total of 34000 Throughput.
Now we can define the ratio that we will use to convert each hero’s Throughput to fantasy points. For this exercise, let’s say that we can consider a “good” player performance to generate about 50 fantasy points from their Throughput. Working backwards:
For Reinhardt: 50 fantasy points = Scaling Factor * 34000 Throughput, so his Scaling Factor will be roughly 1⁄700.
For Soldier 76: 50 fantasy points = Scaling Factor * 21000 Throughput, so his Scaling Factor will be roughly 1⁄420.
Since 1⁄700 and 1⁄420 are very small numbers, it is easier to describe the conversion in the way Fantasy Football does instead: Soldier 76 players will generate 1 fantasy per 420 Throughput, and Reinhardt players will generate 1 fantasy point per 700 Throughput. This process is simple to understand, and can be applied for all heroes in Overwatch. Furthermore, this automatically addresses hero swaps: if someone were to swap from Soldier 76 to Reinhardt mid match, the scaling factors would assure their Throughput would still be representative of their impact.
PPR vs Skill Shots
In fantasy football, some leagues choose to reward wide receivers with additional points in the form of points-per-reception. For Overwatch fantasy, a similar statistic can be created for each hero that I will refer to as “Skill Shots”. Every hero in Overwatch has abilities or mechanics that make them unique, and the usage of these abilities could be translated into fantasy points. This statistic should highlight each hero’s strengths and be representative of skilled usage of that hero.
What would these Skill Shots be? Let’s theorycraft examples. First, some simple ones:
Reinhardt: Firestrike hits
Ana: Sleep Dart hits
Soldier 76: Helix Rocket direct hits
McCree: Flashbang hits, Headshots
Roadhog: Hook hits
Doomfist: Wall-smash hits
Sombra: Enemies hacked (non EMP)
Mei: Enemies Frozen
Junkrat: Enemies Trapped
Then there are some heroes that are more difficult to think of a proper skill shot statistic to track, like:
Genji: Deflect …eliminations? Dash resets? Shuriken headshots?
Tracer: Stickies attached?
Pharah: Rocket direct hits?
Mercy: Players Resurrected?
Lucio: Knockback kills?
Hanzo: Scatter Arrow eliminations?
Symmetra: Turrets created?
Bastion: Recon Mode eliminations?
This fantasy statistic will need creative solutions to apply to heroes like Bastion and the non-Ana support heroes. However, it is important to provide a more variable statistic for fantasy owners to debate. “Do I draft aKm who has more consistent damage output but less Helix direct hits than Tviq? Do I draft SoOn because he lands the most stickies, instead of Grimreality who has higher overall damage?” These are the kind of questions the Overwatch fantasy league should prompt its users to consider.
Kills and Deaths
One statistic that video games can track that football (thankfully) does not is how often players kill each other and how often they die. That would be…somewhat awkward to track in real life. Regardless, it would be a shame not to include the flashiest of all multiplayer video game statistics, K/D, into fantasy Overwatch. Although for Overwatch, we would use eliminations rather than kills – it is a team game after all. Judging from Overbuff’s heroes page, tanks and DPS alike tend to reach a little over 30 eliminations and a little under 10 deaths at the 90th percentile of Overwatch players. However, supports need a bit of help from their assists stat to reach DPS/tank numbers. Luckily, adding both offense and defensive assists at half of their value to their eliminations brings most of the supports pretty close. Here’s a chart of a couple DPS, tanks, and all of the supports to demonstrate a point breakdown of 1 fantasy point per elimination, -1 fantasy point per death, and 0.5 points per defensive and offensive assist:
We see that due to his hero design, Zenyatta breaks this system. It is too easy to get assists with Zenyatta because of his Orbs, so let’s tune his assists down to 0.25 instead of 0.5:
That’s a bit better. Similarly to Throughput, eliminations, deaths, and assists may need scaling like the Zenyatta example above to make it feel like you’re truly drafting the player, not the hero.
…Or not. Part of the fun in fantasy leagues making the decision to draft Boombox over unKOE, knowing that Zenyatta is a great fantasy point-generating hero and because you know that Boombox is likely to play much more Zenyatta than unKOE:
Looking back to Tviq’s Terrors vs Reinforce’s Ringers, RR decided to draft two Soldier-heavy players in aKm and Babybay, knowing that their Helix Directs would garner them a lot of fantasy points. However, TT’s decision to draft two Zenyatta/Ana players won them the matchup because they had faith Zenyatta’s ridiculous assist rate would carry them. This depth of strategy is common in traditional fantasy leagues and should translate well to Overwatch.
Point Captures as Multipliers
As I was theory-crafting how to convert Point Captures into fantasy points, it quickly spun out of control in complexity. Fantasy point should be simple, so why not just use Point Captures as a multiplier to normalize matches of varying duration? Maps in Overwatch generally fluctuate around four* total points played by each side combined (stats from 1 year + of pro matches):
I propose all other stats be scaled as if the map had had four points played whenever less than four points are played. For example, if Rogue defeated CLG 1-0, this means each team’s fantasy points accumulated would be doubled – including the negative ones – since only two points total were played. If Rogue were to play EnVyUs in a 4-3 nail-biter, this rule would not apply, since seven total points were played.
This should help stomps still feel like stomps, since teams that fail to cap a single point will undoubtedly have much worse stats relative to their opponents. It will also ensure that the defending team will not be “punished” for being good at defending.
*Note: Control maps are depressed in their total points because Apex has traditionally played these as Bo3s rather than Bo5s. The Overwatch World Cup has also moved to a Bo3 system, so Control maps may need their own unique normalizing factor (probably around three points). I have no idea why Dorado is so hard to attack, but apparently it is.
Other Potential Fantasy Stats
Rather than continue to pile on complexity, I’m going to list – but not flesh out – a few statistics that could be used as other fantasy point sources, or even as “bonus stats” to add a little flair to individual leagues if their commissioners choose.
Ultimate-based fantasy points: These would work similarly to Skill Shots. Ultimates are game-changers and Ultimate eliminations could function as a touchdown-like mechanic for non-Support heroes.
Combo/spree-based bonuses: What’s more hype than kills from Ultimates? Wombo combos! Leagues could add a Combo Kill / Team Wipe fantasy point boost to award players for participating in Overwatch’s most hype moments
Melee kills: Melee kills are infrequent but still interesting. They could be added as a bonus “fun” stat.
Threshold bonuses: Above, I mentioned that the best Soldier 76’s average roughly 21000 damage per game, but what if a player significantly exceeded that number? Some fantasy leagues grant bonus points based on reaching a threshold, which can make your team’s stars feel more like the heavyweights you thought they would be when you drafted them.
Other negative stats: Did you tank Earthshatter into a Shield? Or launch their Graviton Surge into a Defense Matrix? One of the more entertaining things to do in fantasy is to assign huge amounts of negative points to silly things – fantasy giveth, fantasy taketh away.
For the Overwatch Fantasy League to succeed, it will need one main thing out of the Overwatch League: consistency. When drafting players, fantasy owners need to be assured that their players will be playing often enough to start them in their starting lineups, or they will have to draft 12+ player lineups to have backups to fill in depending on OWL’s schedule.
There also needs to be some sort of structure to the maps played for fantasy scores to be consistent. Assault maps will generate vastly different point profiles than Control, which will have very different outcomes than payload/escort maps, so fantasy owners must have some guarantee that OWL matchups will include similar map type pools across the league. This could be solved by continuing to require that one of each map type be played (the Apex system) per Bo5, and only counting those three guaranteed maps towards fantasy point totals. This means that ideally, the Overwatch Fantasy League would have matches played between every team on a weekly basis with a guaranteed map pool including one assault map, one payload/escort, and one control map.
However, these issues will likely be addressed by Blizzard anyway if they truly are already working on an Overwatch fantasy system. If I can create a theoretical OWL fantasy league on my own, I have nothing to worry about on Blizzard’s end. Rest assured, I’ll be spending my time studying Overwatch player stats and preparing for the draft as soon as the fantasy system is announced.
Until next time,