War Robots WR Data and Analysis

By the Numbers: Doom Blade versus Counterspell

In this week's analytics thread, I've moved to High Bronze- and have some close encounters of the MagGep kind

So there’s this really interesting principle I’ve come across in the design of Magic: the Gathering that basically goes, we think it’s the what, but sometimes it’s the how. By nature we tend to talk a lot about things in terms of outcomes and results, but particularly in the world of games we ignore the experience along the way at our peril. Let me explain.

Imagine for a moment that you and I are playing a game of Magic: the Gathering. Don’t worry, you actually don’t need to be a Magic player to sort of follow along.

It’s a tense game, lots of back and forth. Or maybe it isn’t, and we’ve just been waiting for the other to move. Doesn’t matter. What does matter is that I’ve just drawn my best creature, and I’ve got the mana (resources) to pay for it. I look at the card again, trying not to smirk. I look over at you, over at your board. You’ve got three cards in hand, and two untapped lands in front of you.

I’m smiling now as I tap the lands needed to summon my creature, eager to put the game away. What happens next reveals a fascinating insight into gamer psychology, as one of two nearly identical but very distinct things transpire. 

Possibility A: You tap two lands and play Counterspell. My creature fails to resolve, and gets placed into the graveyard.

Possibility B: You do nothing, and my creature resolves. I can’t wait to attack with it next turn, after the summoning sickness wears off. Alas, it’s never to be- at the end of my turn, you tap two lands and play Doom Blade. My creature is killed, and gets placed into the graveyard.

On the face of it, pretty much the same thing happened. I was excited to play the creature, but was thwarted. In foiling my plans, you used the exact same resources (two lands), and it ultimately had the identical effect (my creature was placed into the graveyard). Both outcomes are basically the same, right? And yet, research from Wizards of the Coast has conclusively shown that the difference in these two outcomes is vast.

So what gives- and how is this related to War Robots? Great question. This being the weekly data thread, let’s do the data first.


It’s been a fascinating and eventful holiday week on War Robots for me. For one thing, late on Christmas Eve my little “fetcher bot” which goes and gets me 60 Gold each night came back with the latest haul. That put me at 4,996 Au, just shy of the fifth Hangar slot I’d been saving for. One quick match later, and the coveted slot was mine! You might recall last week I temporarily replaced my MagGep with a Magnum Destrier. I didn’t want to change that, but thought I’d give myself an “ace in the sleeve” option in case I ran into clubbers (much more on this to come).

Since I only own three Magnums, I tried sticking the third Magnum and two Pinatas on the recalled Gepard, but that seemed to be a muddle. With Pinatas, you want to jump out, ambush, and get back behind cover while waiting for the reload. Magnums need you in the mix. I quickly dropped the singleton Magnum and went full Pinata, just as it was on the day I acquired it.

The other big change since last week’s stats post was that I added a “death button” Golem to the lineup (replacing the Orkan Cossack), which I discussed in my most recent Battle Stories post. This seemed to nudge me up a little in the matchmaker’s eyes, and it’s certainly felt like I’m up against a lot more Medium bots than I have been in the past. Still no Yamantau, though, so maybe this is what High Bronze looks like?


So I collected another round of data. A slightly larger sample size this time, 132 data points instead of 106. As before, I wrote down the pilot’s level, trophies, hangar slots, bots, bot levels, and highest weapon level on each bot. Here’s what I found, including the comparison to last week’s values.

Average Player Level: 15.61 (+2.37)

I’m not sure what I was expecting, but it was probably more than an increase of 2.37 over last week. Of course, we often come across new players’ tales of woe over on the wonderful Wiki Forums, where they simply bought bigger bots and/or leveled things too fast before they learned better, and got bumped up by the unsympathetic matchmaker.

Sure, you can pilot that, son, says the matchmaker. You can pilot that right into some veteran’s kill tally. Have fun in Gold!

Now Level 27, I’m leveling up a lot slower than I was before. That’s perfectly fine- even expected- but it does mean that my experience level is still significantly higher than many of the opponents I face on the field of battle.

Average Trophies: 141.80 (+109.88) – Median Trophies: 26 (+9.50)

Gott im Himmel, what a difference! This is the single biggest marker I found that statistically validated my feelings of being in a stronger pool than in the past (again, as a likely result of the Golem). Just like last week, though, we need to unpack these numbers a little bit to understand them.

At the end of a long Christmas Day, which included not one but two full Christmases (gift exchanges, food, etc) at two different family homes, I was delighted to carve out a little unwinding time for myself. I cracked a beer and fired up War Robots.

And ran straight into clubbers.


As discussed in last week’s stat’s piece, these have been akin to shark attacks- generally unpleasant but specifically rare. No big deal, I thought, picking myself up off the ground. Time to go again! A few rounds later, I did something new: I put the game down.


Simply put, the game wasn’t fun to play. I don’t know what it was about the Christmas spirit that prompted so many clubbers to come out of the woodwork to harvest the rest of the playerbase. Or perhaps this was random chance that put me in front of so many trash guild reapers.

Wiki Forum contributor sweetpickle had made this observation in a response to last week’s post, which also seems applicable.


In any event, the uptick evident in this week’s trophy average is entirely attributable to the clubbers, whose average trophy count was a staggering 1,136, operating in an environment where the average (non-clubber) trophy count was 41.69. If you’re an anti-clubber and looking for a statistical nugget from this week’s episode to say “see, there IS a problem,” that’s probably your huckleberry.

For the interests of transparency, I’ll remind those keeping score at home that my trophy count stays fairly consistent at a little over 300. If I were trying to make a direct case against clubbing, I’d be open to the characterization of, “he just wants the clubbers to go away so he can club them with a slightly smaller stick.”

Fortunately, I prefer a more nuanced middle ground, one that accepts both that clubbing is a problem in the game, but one that has a reasonably low occurrence. This is highlighted in the difference between the mean (where outliers can have a distortionary effect) and the median (which often gives a clearer picture of “how things really are”). The median trophies were up significantly, from 16.5 last week to 26 this week. That certainly helps substantiate that I’m now in a more competitive bracket.


Percent of Players with Three Hangar Slots: 73% (-21%) 

I continue to be surprised by the number of players who haven’t appeared to make a fourth slot a priority. At 5,000 Au, the fifth slot is a real grind, but the fourth seems relatively affordable and adds is a significant upgrade from three. It’s not just an “extra life,” it’s also the ability to have a greater variety of bot/array to give you the strategic flexibility to respond to any situation you find yourself in.

Just under 9% of the players catalogued this time around had the full five slots in their Hangar, and all but two of these were clubbers. One player ran a field of five Thunder Schutzes, which I don’t designate as clubbing but has increasingly been noted as a viable Gold-farming vehicle. A special shout-out goes to the player who had five slots, but was running a single Gepard, three Cossacks, and a Destrier. Sir or madam, whoever you were, I salute you.

By the way, just under 6% of players were running only two slots. Hangars first, peoples!

Average Robot Level: 2.27 (+0.17)

I don’t expect to see a lot of change here. This really took me aback last week, but I we still have some crossbreezes intersecting here between “I’m still learning to play the game,” and “I want to stay in Bronze.” Players just aren’t going to be jacking these levels up, because once they’ll essentially self-exclude from my calculations. I’m staying in Bronze.

For you visual learners

Average Highest Weapon Level: 4.50 (+0.64)

This tells me folks are taking a little more care of their moneymakers, but this is subject to the same soft-ceiling considerations as bot level. On the whole, 4.50 is probably pretty good.

Percentage of Clubbers in Population: 9.09% (+8.15%)

That’s a huge increase from last week. Put another way, I was ten times more likely to run into a clubbing player this week than I was last. This is concerning, because you’re about to enter a game with eleven strangers. Of course, five of them are on your side, and while a plausible case can be made that even being teamed up with a clubber takes some of the fun and challenge out of the game (which I’ve found), that’s not quite the same thing as being griefed.

So maybe our chance of running into a Red clubber is closer to 5% than 9.09%. Even then, I’d argue, it’s smaller still. After all, while one clubber in a game is bad enough, these folks usually enjoy traveling in packs. Sometimes you’ll get two. Once in awhile perhaps three. So just spitballing numbers here, but taking ‘clustering’ into account maybe the actual percent chance I’ll be battling clubbers is around 3%?

And- if you’re looking for an upside in all of this- the good news is that if you’re up against multiple clubbers, the game is probably going to be over fairly quickly and you can move on to another.

So it seems that even with my run of Christmas luck and the higher prevalence of clubbers overall this week, perhaps the “shark attack” theory still (forgive me) holds water?


What Bots are People Running?

This is probably my favorite part. It’s been interesting seeing more of the Medium bots at work, learning how to counter them and facing a greater variety of challenges. Here’s how the list breaks down, with anything above a 2% change highlighted.


I don’t think there’s a lot to be surprised about in there. Destriers are the biggest loser, which is exactly what you’d expect from everybody’s “starter bot.” While I’m using mine as a MagGep substitute (same Magnums, less guilt!), they’re an easy drop- especially for those with a small allocation of hangar space. I was sad to see the Cossacks drop down a bit, but I’m suspecting that they’re increasingly a niche bot as players advance. Much like vintage Volkswagens, folks who love them really love them, but most aren’t all that interested.

The big winners? Surprise! Gepards, or course. With more advanced play at this level (and, yes, more clubbers too), they’re a no-brainer choice given their combination of high speed and survivability. Having added a fifth slot this week, my Pinata Gepard is back in action and I’d forgotten just how much fun Gepards are to play.

And now I can also stop agonizing over a Gep-less Hangar with a blog called Gepard Diary.

It’s the little things, innit?


I started this week’s update with a discussion about Magic: the Gathering, and two spells in particular: Doom Blade and Counterspell. One of them kills a creature, the other one counters a spell as its cast. In both cases, the effect is about the same. A creature card ends up in the graveyard, having done nothing in the game. In both cases, the offending player has used the exact same number of resources: one card (the counter or removal), and two lands to pay for it.

But as mentioned above, Wizards of the Coast’s player research has found that the emotional response (of the player who lost their creature) to both scenarios differs widely. So much so, that it’s altered the course of the game’s development. Countermagic has been dialed way back over the years, while creature removal may fluctuate a bit in efficiency but is far more widespread. Why?

As it turns out, a player who at least gets to add their creature to the battlefield gets a certain satisfaction for having resolved the spell. Everyone knows creatures can be killed, so losing it that quickly sucks, but when it’s your time, it’s your time.

Having the same creature countered, however, feels much more like your opponent didn’t even let you play the game. There’s a reason countermagic is commonly referred to as “permission,” and when permission is denied, it feels much different than when the normal course of the game happens to not work out in your favor. Permission denial, Wizards discovered, is much more of a “feel-bad moment.”

I’ve come to see this dichotomy, as it relates to War Robots, because of my experiences on Christmas. If you ask any player if they like losing, they’ll naturally tell you “no.” There are plenty of skilled players out there who could crush your dreams with just about any bot, but somehow if they do so in, say, a Taran Cossack, it’s different than a Magnum Gepard.

That’s probably because to many, one of these let you feel like you had a chance. And the other one, well, it kinda didn’t- even though the odds of you surviving the engagement might well have been the same. There’s a lot of aspects to my game play that need some work, but I can tell you that most of the time now when I catch a lone beacon runner in the early part of the game with my Taran Cossack, I’m going to sky them and take that beacon. I’m not piloting a MagGep– but I might as well be, for all their chances of survival.

But I got ’em with Doom Blade, not Counterspell.

So as mentioned at the outset, we often think it’s the what, but it’s really often the how. Some dismiss the frustration of running up against trash-guild clubber-farmers as, “well, they just don’t like to lose.” Maybe there’s more to it than that.

As we continue to have the ongoing discussion about clubbing and the overall health of the game, the data I record is going to help put at least some context around prevalence. But by the same token, it’s important to remember to consider the perspective of the playerbase- even if it isn’t entirely reasonable, on the basis of actual fact. We play games for entertainment, and the idea of “feelings” has a very important seat at the entertainment table. People don’t play games they don’t enjoy.

Anyone simply saying “clubbers aren’t really a problem at all, just look at Dredd’s data” isn’t seeing the entire picture. Just as someone saying, “clubbers are ruining this game.”

The truth lies somewhere in the middle. As with most things.

If you’ve gotten this far, thanks for reading!

6 comments on “By the Numbers: Doom Blade versus Counterspell

  1. Super insightful as always! Your blog posts really get me thinking about the game from different angles 🙂

    Liked by 1 person

    • Thank you for the comment, Michael! It’s always interesting to me how many ways there are to approach this game, it has a delightful amount of depth.


  2. Pingback: Battle Stories: Exit Stage Bronze – Gepard Diary

  3. I am enjoying reading your blog. As one of “those” players with a gepard hanger, I am one who others would probably consider a clubber. I am part of a top 50 clan on IOS with a consistent win rate of 90%, usually 1500+ cups (I play a lot) and almost 6000 total victories. Not trying to brag, just want you to understand my perspective.

    I often run in 3 tiers – high bronze, silver and low gold. I am going to share my hangers and tiers so you and your readers can try them out. I will start with my silver hanger, because everything is just a slight modification from that.

    For Silver:
    I run 5 gepards, all level 3-4 bots, and then max level weapons (12). 3 with all ep mags, 1 aphids, and then 1 either geckos, spirals or crv pins depending on whether I am solo or with my squad.

    For Bronze:
    Same gepards (level 3-4), but now only one level 12 weapon, and the other two are level 5-6. Then I usually run 4 mag geps, and one aphid.

    For low gold:
    Same gepards, but only 4 of them. The 5th spot becomes a level 4 Galahad with one level 12 Taran or Orkan, and 2 level 6 aphids.

    I offer you this information because I prefer to face opponents based on skill and not a bot advantage. Anyone can setup their bots like I do. I welcome the challenge.

    Hope to see you out there!

    P.S. – I have lots more information to share – feel free to contact me if you want to start a dialogue.



    • Hey Falcon, great comment. I’d love to develop this into a full article. Would you mind shooting me an email at GepardDiary (at) gmail?


  4. Pingback: By the Numbers: The New Matchmaker – Gepard Diary

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: