Balancing based on the Smite Pro League

With the return of the Smite Pro League, we have the return of an interesting question: How much should game balance be based on the SPL? On the surface it may seem simple. As the highest level of play game balance should be mostly dictated by the SPL as the pros know the game the best. While I’m not denying the pros' knowledge and experience, there are some aspects about the SPL that make it not suitable for fully basing the game balance on. When looking at god balance, you have to choose the most reliable stats from the correct places.

First, let’s look at the aspects of the SPL which make it problematic for meta analysis. At its core, when it comes to statistics, the SPL has one fatal flaw: it’s a small population. Even if the SPL has around 12 games per week for anywhere up to 10 weeks, in comparison to the general Smite population, this is a small sample size. This means that win rate statistics are very easily skewed. Another thing that contributes to the unreliable win rates is how volatile these games are. At the top level of play, the tiniest thing can cause a change in how the game is going. This means that wins are never set in stone, even if you play your best. Just recently, one of the bottom seed teams, Cringe Crew, took a game from Panthera, the then undefeated top seed in Europe. The SPL is unpredictable and without enough games to make these sort of situations outliers, it affects the win/losses drastically.

So what can we use from the SPL? Well one stat that does have a large enough population for it to be relevant is the Pick/Ban rate of characters. This effectively has double the population because it includes bans as well as picked gods. Furthermore, pick/ban rate is not affected by how good the teams are, taking out the volatility aspect. The only thing that can skew this ratio is teams doing specific pocket strategies such as picking non meta gods. Thankfully, this happens rarely enough that it can be ignored. Therefore, in my examples, I will be using the pick/ban rate of gods in the SPL to talk about their balance.

But we also have to talk about win rate somehow, as two characters that both get picked and banned a lot need another way of being compared. Therefore, we will be looking at win rate on PC Conquest Ranked as our second statistic for balance. These statistics will come from http://smite.guru/gods.

Let’s look at our first example – Raijin. Raijin is infamous at this point for having a dreadful SPL win/loss ratio – 1/10 at the time of writing. Looking at this, this means that Raijin only wins 10% of the time right? Well no, not really. If we go to SmiteGuru.com, we will see that Raijin’s ranked win/loss is 46.17%. And while this falls short of the desired 50%, its closer than 10. What about Raijin’s Pick/Ban rate in the SPL? So far, its 47.5%, surprisingly close to his ranked win rate. What does this mean? It means that Raijin is in a good spot currently, with a win rate close to 50% and a pick/ban rate that is very similar. If Raijin was buffed based on his 10% ranked win/loss, it would turn a balanced god into an overpowered god.

It should also be mentioned that Raijin is a new god, decreasing the sample size for win rate even more.

So why is Raijin’s SPL Win/Loss so shockingly bad? Because of the small population size mixed with the volatility of SPL games as we mentioned above. The ranked win loss is based on thousands more games, which allows for the volatile and throw games to become outliers, rather than a large part of his statistics. If Raijin was buffed based on his ranked win/loss, he would become downright overpowered.

Let’s look at another god that the community demands be buffed: Arachne. Arachne’s SPL Win/loss is not applicable because her pick/ban rate is 0%. Arachne has not been picked once in the entirety of the 3 or so weeks we have had the SPL running. So we would expect that her ranked win/loss is dreadful right?

Arachne has been virtually unheard of in the SPL for a long time

Arachne currently stands at a ranked win/loss of 58.9%. That win rate correlates to a god slightly over the point of balance. So why aren’t we seeing her in the SPL? With a high win/rate in ranked and an extremely low SPL Pick/Ban rate, Arachne suffers from 1 thing: Power creep. Arachne is a good god, good at what she does, but there are other gods that do it better, therefore making her irrelevant to the pros. A situation like this cannot be fixed by simple stat buffs. With such a high win rate already, stat buffs would just make her OP. Arachne needs to have her kit altered, to give her a niche and make her relevant to the SPL.

A controversial buff was recently given to Serqet, with many people complaining that HiRez just wants to see the same gods in SPL games. Is this true? Well, in terms of SPL, Serqet’s pick/ban rate is 38.8%, which certainly isn’t awful, but it’s a definite fall from grace from season 2 Serqet, the then Queen of the jungle. What about in ranked?

Much to my own surprise, Serqet is currently the worst Assassin in ranked, with a win rate of 43%. As silly as it may seem, Serqet may have needed that small buff. This is also corroborated by the fact that we are no longer in a crit meta, whereas Season 2 was very much a crit meta. Serqet previously needed crit to be effective within the game, and with crit’s downfall, her effectiveness fell too.

So what have we seen so far? We’ve seen that the SPL is not a viable source for telling if gods are balanced because of it’s the way it’s played, and that SPL stats can often be completely wrong due to phenomena like power creep. But there is one god that breaks our nice set of data. One god that totally goes against everything we’ve said.

Khepri’s introduction changed how the game of smite was played. Khepri’s ult is such a departure that he really can’t be classified as a support or a guardian, he’s just Khepri. Khepri boasts a SPL Pick/Ban rate of 95%. Because of the way Khepri is, he’s a simple must have or must NOT have.

Khepri has been a force to be reckoned with since his release

For readers that don’t know, Khepri is a god with an ultimate that when cast upon an ally or himself gives them multiple buffs. Furthermore, if the target dies during the duration, they resurrect next to Khepri with around 40% of their health bar, ready to continue fighting. This ability can change a team fight, and makes Khepri an incredibly safe laner.

Khepri's Ult as of April 15th 2016 - Screenshot from http://smite.gamepedia.com/Khepri

With the high SPL Pick/Ban rate, you’d expect Khepri’s ranked win rate to be hilariously good, but it’s not. 54% is a good win rate, but it doesn’t match up to his SPL pick ban rate. Why? Well when we look at Khepri, his combat resurrection is powerful, but to turn it into a game winning play, you need the team co-ordination that only a full team on comms can provide. Don’t get me wrong 54% is good, but it doesn’t match up like Raijin’s statistics do. So did Khepri deserve his endless nerfs?

The difficult thing to realize is, as much as the pros play the game different to us, they are still playing the same game. This means it has to be balanced for them as well. This is why Khepri got obliterated by nerf after nerf after nerf, because he was still downright overpowered if played on a full communicating team.

So what we’ve seen is that looking at SPL stats themselves is very misleading. In fact, balancing based on them could very well lead to an unbalanced game overall. So why does the community still ask for Raijin buffs and Arachne nerfs? Simply put, the SPL is the most widely accessible way of watching the meta in action, even if it’s a very small scope with the small amount of games played. But at the same time, it has to be considered because of the fact that SPL often shapes how the meta is formed in terms of when and how often gods are picked. But to look at how successful this meta is, you need to look at Ranked win/loss rates.

So next time you go to complain that your favourite god is underpowered because of their SPL winrate, think about what we talked about in this article. Sometimes, you just have to look a little bit deeper to really see meaningful data.

Get your own AKRacing Chair here and support our players, all profit goes towards the teams!