The Most Important Statistic in Gears of War

I’m a big fan of statistics. During some of our early tournaments I was known as the guy who brought a clipboard and notebook with him to events and would take down the stats after each map we played. I felt that we could use these stats – which I had broken down by map and player – to improve our chances of winning. The metrics I tracked were the same which the game presents during a match:

  • Kills
  • Downs
  • Deaths
  • Revives
  • Points
  • Rounds Won
  • Rounds Lost

After an event I’d spend the couple hours waiting for my flight back home compiling the numbers to see how we did individually and which maps gave us the most issues. I did this a couple major events before I came to a conclusion: You can’t really determine anything from looking at these stats in Gears of War.


Kills and downs are entirely dependent on what position you play on the map and who is closest to the kill since “kill stealing” isn’t something which is a factor when you play to win. Deaths were irrelevant since a lot of the time a death isn’t your fault – one of your team dies and you all fall in sequence. More on that shortly. Revives are a pointless thing to track since they are driven by opportunity much as Kills/Downs are.


Points are a pretty solid indication of how well you’ve done, but they tend to go the way of players who grab power weapons a lot. A player with the Longshot in Gears 1 could get 3-4 meaningless downs in quick sequence and rack up 20 points doing so. Likewise, some positions require simply holding down an area rather than stopping pushes or killing the other team. It’s probably the best stat of the base group for determining effectiveness, but there are so many anomalies that comparing players based on points over the course of an event is pointless.


Rounds won and lost are good tools but host advantage in Gears often dictated you’d lose maps based solely around that factor. I could look at the rounds won and see we went 1-3 over the course of an event on War Machine, but in reality if we played all four of those maps off host then that’s not a bad record in competitive play.


I feel competitive play is the ideal place to look at statistics because it eliminates a lot of the potential variables present in public match play. In a competitive match you have similarly skilled teams who are guaranteed to have four players communicating together and have an extensive knowledge of how the game plays. The odds of getting a public match where all 10 players have microphones and are playing to win is astronomical.


Are stats from competitive play entirely indicative of how they would look in public play? Nope. But I wager that the better the overall community gets, the closer they get to the numbers I’ll be presenting below. As weaker players leave the game for greener pastures, the “hardcore” community is often all that is left and the level of play in your average Gears room improves. I think we all saw this during the beta with each wave of new players that came in – you’d see a skill dip and then a rise each time a group was added and quit playing.


The other advantage in looking at competitive statistics is simply that I have access to them – several teams have been uploading videos of scrimmages and matches over the past few years and they are pretty easy to find on Youtube. Searching for public Execution matches proved pretty futile – all anyone uploads are games where they go nuts and those are poor for gathering accurate statistics. Most public matches uploaded to Youtube are anomalies or extreme mismatches. As such, I’ve focused on competitive Execution, particularly MLG settings which are as follows:


  • Execution 4v4, First to 4 Rounds
  • 4 Minute Rounds
  • Maps: Avalanche, Blood Drive, Jacinto, Pavilion, River, Security
  • Lancer Starts
  • Most maps feature one or two of the following: Boom, Torque, Sniper
  • These weapons are swapped out: Frags, Mortar, Scorcher, Boomshield, Mulcher, Hammer of Dawn


With that, I think I’ve explained enough to start getting into some numbers.




A typical CS 1.6 match page on Gotfrag's Gamesense

Several years ago I came across an interesting article about competitive Counter Strike. It interested me because CS1.6 plays out similarly to Gears of War Execution – each team starts with one life and has to eliminate the other team. There are differences, of course, but the core idea is the same and so few games have a 1-life mechanic that it made me interested in the results (and I also enjoy watching CS1.6 matches).


GameSensed: The Importance of the First Frag


GameSense was a stat tracking software which ran alongside the game and reported stats from matches live and compiled them into all sorts of formulas. It tracked just about everything, including a ton of things the base game didn’t. All of this got instantly updated to a database of sorts which kept track of player and team performances and allowed for tons of analyzing of playstyles and such. But the important stat turned out to be one which I’d suspected was also important for Gears – who is getting the first kill?


CounterStrike 1.6:

% of Times Winning the Round When Getting the First X Frags

First X Frags

Round Win %


74.8 %


90.7 %


97.8 %


99.8 %


100.0 %


To make that read clearly – if a team got the first kill then died shortly after to even it up to 4v4, they still won 75% of the time. If they got the first two kills before dying, they won 90.7% of the time and so on. Only once in every 500 rounds did a player win a 1v5 scenario.


The thing that is significant about this is that CS1.6 has far faster kill times than Gears of War. This allows a team which is down a man the opportunity to quickly even the score with some swift headshots. Despite that, a comeback from 3v5 only happened 9.3% of the time. Teams are able to take advantage of being a man up and would systematically work the map so that they could complete their objective by arming a bomb or killing the other team.


So how would this look in Gears of War – a game with longer killtimes on the whole, but also featuring strong power weapons on the map and the lack of an objective?


Obviously the game doesn’t keep this stat (it should!), so I went and scrubbed through 19 maps from MLG scrimmages and matches that I could find on Youtube. When possible, I preferred entire sets so a full match was presented rather than any anomalies from people going off during one map. I’ll list all of the matches in an appendix at the end for anyone who wanted to take a look. I tracked 105 rounds in total. Here’s the results:


Gears of War 2:

% of Times Winning the Round When Getting the First X Kills

First X Frags

Round Win %









(Note: I went about tracking this for the few non-ridiculous scoreline public matches I could find on Youtube – I tracked 6 maps and the numbers came out almost identical to those above, with even fewer one-man-down wins. I couldn’t find enough for proper statistical analysis.)


The 2v4 only occurred once in 105 rounds, and it was a bit of a fluke – 2 players died for each team right after each other which made it a 2v2 scenario. In 57 attempts, that was the only time a team managed to win at a disadvantage.


The 1-man-up scenario is a bit misleading as well. Just like in Counterstrike, during a competitive match there is very often a push which results in a player dying from both teams. It isn’t uncommon for a team to push and kill one player and simultaneously have one die. I recorded the numbers very strictly and didn’t give leeway there – but it’s worth noting that roughly half of the 3v4 wins were scenarios where it was instantly evened out to a 3v3 scenario.


By scoring the first kill you give your team roughly a 75% chance of winning, and getting the first two kills means your team will win unless you make some incredible blunder. I believe this conclusively makes “First Blood %age” the most important stat in Gears of War in both Warzone and Execution.


Mini-Rant: Why you shouldn’t care about “Kill Stealing”

Both kills and downs are largely meaningless statistics. The amount of times that you do all of the damage to an enemy and someone steals it are far outweighed by the amount of times two players fire at an enemy and only one down and kill are distributed. You’ll get downs you may not deserve, as well as kills you don’t deserve. Just because you got the down doesn’t mean you deserved the kill either.

Working towards an impressive Kill/Death will often mean you had to step over a few heads and undoubtably get many kills which were really team efforts. You can choose to be proud of your K/D and feel it makes you a great player, but the truth is that it’s heavily influenced by those you’ve played with. Likewise, the playstyle you employ to get that K/D may not really be putting your team at an advantage – just above I noted that the first kill of a round has the single biggest impact on a match. How often are you going to get that first kill if you’re protecting your K/D? The more likely scenario is that you’ll be faced with a 1v4 and get 1-2 kills with grenades and take solace in the fact your K/D is in tact while still losing the round due to having a .5% chance of winning.




Cole celebrates a steep increase in odds for this round!

So why is First Blood so important? After all, it’s only one kill, and a team just has to get one back to even the score. There are a lot of factors, but I’ve listed some below in order of importance (in my opinion):


  1. Map Control – The single biggest factor in winning a round in Gears of War is map control. Generally, off the start of a round teams will push out for power weapons with more emphasis on the best weapons. A kill occurring can mean that a team just lost a power weapon, but even if that isn’t the case, they are now less capable of contesting a weapon. They also are more susceptible to crossfire and 2v1 pushes. Our team had a call we used called “RESET!” which meant for us to run back to our end of the map after a push where we got a kill or a weapon. This meant we were in control of the map from there on out and could work our advantage.
  2. Stopping Power – This was added in GoW2 and I believe it had a noticeable impact on the amount of clutch plays (along with the move speed reduction). To recover from being down a player, you need swift kills. Firing a rifle at a smart player is going to result in them rolling to cover more often than not, meaning the best chance of getting a kill is to push with shotgun. Stopping power, combined with open maps, means that a team can keep someone at bay pretty easily and eliminate the possibility of a shotgun kill.
  3. Lack of strong “freebie” weapons for both teams – In GoW1 the most common times I’d see clutch plays was when a player had a sniper on Fuel Depot or Canals. Since each team got one on their side of the map, it was usually quite accessible for a team that was down a man. Active downs made it very lethal in a good players hands. In GoW2 the only strong neutral weapon was frag grenades, which promote a very passive style of play. That passive style of play also means that the team with more players can take their time trying to pick someone off. I believe the new stun Ink Grenades and insta-kill Incendiaries will help fix this issue in GoW3 since both can be used offensively to even the score. Really looking forward to the first 1v2 clutch I get with Incendiaries.
  4. Third Person – This can’t be helped, but it has a huge effect on clutching in Gears of War. Being able to hide behind a wall and see someone coming is a bonus to a defender – they get to see their prey coming towards them and decide to back away or lie in wait. More often than not, the defender will get the first shot and cripple someone pushing with a shotgun. When party chat comes into play, they can also relay movements of an enemy well before that enemy can even take a shot.
  5. Linear Map Design – I could list this higher, because on some maps it is a huge factor but on others it isn’t a major issue. Gears has traditionally featured symmetrical maps with 2-3 areas of conflict where the power weapons are placed. This will usually mean that when a team is up a man, they simply need to push forward in a loose line and cover the chokepoints as they push. These points are usually close enough that if a player pushes one direction, the other team can quickly collapse on his location and support the person being pushed. Wider maps such as Mausoleum, War Machine, and Fuel Depot alleviated this issue a bit and afforded ways to move around the map when down a man. Security is a good example of a map which is very difficult to clutch on due to only 2 areas to push across when lasers are up.




An example of some of the stats Gamesense tracked

Since we’ve looked at why First Blood is so important, I want to address some other metrics that could be used to judge performance and the issues they present that keep them from being as important. I really feel like Epic should consider including these statistics eventually given how unimportant many of the currently tracked numbers are, but they don’t quite have the impact of First Blood %age. Some of these are much better suited to respawn gametypes than First Blood %age.


Damage Dealt – This is something which I’ve been requesting that Epic track for quite some time. It’s a great value to keep track of because it shows a players aggression and how much pressure they put on the other team. The “Points” system currently tracks this but then dilutes it with bonus points for kills, headshots, revives, downs, and so on. All of those things aren’t reliable statistics for reasons mentioned in my intro. Damage Dealt falters in a comparative sense in that it punishes players who make strong pushes and die early in rounds. It’s hard to tell a successful pusher from someone who is wildly pushing and dying without getting the kills using just damage as a metric. Regardless, it’s a piece of the puzzle which should be tracked.


Damage Per Round – Just an efficient way of breaking down the “Damage Dealt” stat into something which is meaningful over the course of a lot of matches. It suffers from the same issues as Damage Dealt.


Kills Per Round – Another nice metric which would show how dominant a player is over a long period of time. This is a better statistic than Kill/Death because it acknowledges players who are helping their team finish rounds rather than simply staying alive. It’s another number which becomes flawed when you consider that a player who grabs a power weapon consistently is going to have a higher KPR than someone who does not. Likewise, in a scenario where you have a number advantage it isn’t always ideal to be the one who dives in to go for the final kills. Letting your player with the power weapon take their shots is usually ideal.


2 Kill, 3 Kill, ACE’s Per Round – I’m sure this could be combined into a general “dominance” rating like CS1.6’s Gamesense had. Much like KPR, this would track how often a player has a huge round where KPR is a more general sense. Huge rounds are often done with – you guessed it – power weapons.


Lethality Ratings – This is something which I personally thought of which I’ve tailored specifically to Gears of War. One of the things which I’m really critical of Epic about is how they will often say something like (fictional example) “The Gnasher gets more kills than every rifle combined”. This is usually presented without any context. Gears of War is a complex game and very often the rifles are used to close the gap between players and give a player an advantage as they get to close quarters. That number doesn’t account for how often a player gets a down with the Lancer then finishes with a Gnasher or execution.


Lethality Ratings would be a combination of the following factors: % of all damage dealt with this weapon (roomwide), downs with this weapon, kills resulting from this weapon. That means if you get a down with the Lancer and someone else kills then with their shotgun, the lethality rating would boost for your Lancer. The initial percentage would mean your Lancer is being tracked against everyone else in the servers weapons. It would be an accurate representation of how much you dominated the battle with your weapons. The issue with lethality ratings is the exact opposite of the others – it doesn’t account for power weapon play as much due to limited usage. It would be best served to track rifles and shotguns.


Question of the Day: Is this type of gameplay ideal?


It might feel a bit cheap, but I’m going to end this mammoth leaving you with more questions than answers. One of my goals on the Epic Forums, the blog, and on our Youtube channels is to educate people about the game who may not think about things in-depth or may be new to the game. I feel like the more you know about the game and how it works the better you will be at the game and the more fun you’ll have. Taking analytical approach to the game is one of the ways I’ve found success in tournament play and really started to enjoy the game more. There is some hidden depth to Gears which you only find when you start to think about things in this manner. The slightest change can have a huge ripple effect across the entire game (don’t go stepping on those butterflies!) and not only change how the game plays competitively but also how it plays in the average pub room.


Just recently Epic stated that Execution and Warzone will not have stalemates (but can change that server-side post-release if they want). The team with fewer players would lose the round when time expired. Myself and my teammate arCtiC have covered our thoughts on the subject (Link: ), and hundreds of people on the forums have sounded off their opinions on the matter as well. I’d like for everyone to take a moment to process the information presented above and think about how the rule will impact the game. We can see from the data that a team that is down more than one person will very rarely win a round and the best chance to win a round is to push very quickly after getting the initial death. Man-Up Rule would speed up those 2v4 scenarios but also makes that scenario even more meek for the disadvantaged – to a poiint where we’d see even fewer clutches in that scenario. However, because Man-Up forces quick aggression, we could see an increase in scenarios where a team evens up the count and potentially wins the 3v3 scenario. The most likely result is that the Man-Up Rule squashes that magically rare moment when someone overcomes all odds and wins 1v4 or 1v5. But does that happen often enough to justify drawing out the rounds where a 2v4 loss is practically inevitable?


On another note, what could possibly be done to increase the amount of clutches in Execution/Warzone? I’ve noted what I feel are the main issues, but realistically the only thing listed which can change is stopping power and that would increase the power of shotgun play by quite a bit. Strong rifles are largely to blame for the inability to clutch in Gears, but they also keep the gameplay from being one-dimensional. Increasing the power of weapons that are given to both teams would help, but would also significantly reduce the rifle/shotgun play. I believe an all-around drop in weapon damage would favor the shotgun since getting in close with that will always equate to a kill against a person equipped with a weak rifle. I’m not sure there is a solution for Execution – the core idea of the gametype will always mean that more guns able to focus and control the map will have a significant advantage.


One way I’ve always felt the gametype could be improved is if there was an objective. As unoriginal as it is, having teams plant a bomb (or arm a laser or detonate a wall, etc) and the other team having to disarm it is a solution which just makes sense to me. It replaces the need for the Man-Up Rule to speed up proceedings by offering a way for the team ahead to end the game. It also gives a team that is down a man a slim chance to arm a bomb and hold out just long enough to win the round. Once armed, the bomb serves almost as a player would – demanding the attention of the other team and eventually requiring one player to dedicate the time to disarming it. This certainly wouldn’t fix the problem – as is evident by the numbers presented from Counterstrike – but I think it would make the gametype a bit more dynamic and most importantly, more fun.


But does Execution even need to change? It’s lasted two titles and has been the most played gametype by the Gears populace. That title may be threatened by TDM in GoW3, but I think it will always remain a favorite for hardcore Gears players. I believe it is also responsible for there being such a large skill gap in Gears of War between veterans and new players – Execution/Warzone force you to think about why you just died and then “punishes” you by making you sit out for awhile. It also has you watching players who may be better than you, another way of subtly improving the populace with each and every game. While the action may reach a point where it’s heading towards a predictable end, the tension of fighting against the odds and trying to win that 2v4 or even 1v5 has you on edge and when you finally pull it off it can make your night. Would that euphoria feel as great if it happened nightly?


The Takeaway


You made it! I promised cookies, I deliver.

First off, if you made it this far without skimming, hats off to you. You, Sir or Madam, are a Gentleman or Gentlelady and a scholar. I hope you’ll leave a comment and some thoughts and hopefully check out some of my stuff in the future.


The literal thing I hope everyone takes away from this is the importance of getting the first kill in Execution/Warzone – ideally getting it safely. Hopefully while you play you’ll take note of how often you win when you get the first kill and start to see patterns in why it happens.


The big picture that I hope you take away from this article is that there is more to Gears than is evident on the surface. The stats presented aren’t necessarily the ones you should be worrying about, and there is more to the gameplay than just pointing and shooting at your opponent. If you dig a bit deeper you might just find yourself playing better and finding more enjoyment from a game you had otherwise grown bored of. Maybe, just maybe, I’ve inspired a few of you to take a stab at playing in tournaments for Gears 3.


Appendix with match video links and raw numbers:

coded by nessus

  • Joroth

    Fucking amazing. I’m surprised you would go this in dept in strategy, tactics, and competitive experience to release to the public.  You’re quickly becoming one of my favorite writers. I took my time to read the whole thing I only hope epic does the same. They seem to have a bad habit of refusing to change things they think is right or wrong regardless of what the public wants. (their attitude towards e-sports in the past comes to mind) Now to be on topic I’d love to see some of these stats gets tracked on the leader boards. Personally I always tracked my stats by “points per round” and “points per game” Cause I always hated the K/D system.  You did say you would leave us with more questions than answers, but I want to know from person opinion what stats/plays do you look for in the competitive scene to determine a players Skill/potential? 

  • Jason Walker

    Awesome write up Skyllus, thanks for the cookies. I’m not really liking the man-up rule myself although I hope they’ll add an option to revert back to GoW1/GoW2 style in private (not quite sure if they confirmed whether or not they won’t or haven’t added it). A lot of good points made here, hopefully more people will see this.

  • Nunch

    Nice and very well written article. Had some very good points in there, especially loved the “rant” part. I’ve always been fascinated with all kinds of stats that could be gathered from games, so this gave something to think about.

  • Jason

    That was honestly a hard read but I did it. And I came out with an E-Cookie. I think that is enough lol. Good stuff. 

  • Justin Doverspike

    Really one of the best reads that i think everyone should read throughly. IMO The k/d is something that can be glanced at to make sure someone isn’t horrid (i.e. less then 0.5) but even then shouldn’t be a final judge. I personally look at Points Per Round and Per Game because I think this is a better representation of how well the player is going for the overall win.

    I do believe that they are at least “trying” to do this in Gears 3. I noticed in the Beta that a friend was getting more points being a support player (Staying at high vantage points with Hammerburst calling out people and tagging them) because of the extra points he was getting just for tagging people so we could kill them. There was actually a few games where he beat me in overall points solely because of assists.

    Question back to Skyllus (or anyone who read this): Do you think Epic will try and lessen the value of wanting a kill over the objective in certain gametypes? I know Slayers will always be needed in Annex, Capture the Leader etc.. but would it make sense to try and get even the slayers more involved with the objective to get there points up? or will Epic continue the same formula and you will just have to follow someone and watch them to tell if they are truly good?

    Sry for Essay but thanks again for the read :D   (and the cookies!)

  • Mark Foerster

    I personally wish that the stats would be much more in depth. Lots of times I rack of TONS of damage points and have will have fewer kills than the rest of my team. My K/D doesn’t suffer from it though. I just wish that people sought out more than just a high(er) K/D. If these kinds of stats were implemented into Gears 3 I think it would make “kill stealing” much more a thing of the past.

  • Nick

    Really want to say well done dude. A fantastic read and I honestly feel like I’m taking a lot  away from it. Not only has it made me think about this new statistic, but it also has me wondering about many other aspects of the game.

    Very well thought out and researched. Thank you.

  • Matt Cotter

    You guy’s really stepped it up with this article.  I picked up Gears1 on launch day never once thought about the information you’ve provided in this article in such detail.

    Well done.  As always, I really appreciate what you do for the community.  Keep it up.

  • David Vaughan

    Good Article, Really gave me a good insight to the Statistical side of things which to be honest just about everybody forgets..