Thoughts on RQ Rescaling
I’m working on the balancing of the RQ of cars, and just thought I’d lay out some of the thought processes of checking the ratio between the old and new system, and some other vaguely connected thoughts. This may not be particularly focused and might just be insider baseball, but I’m writing it to help articulate some of my thinking, so I may as well share it.
This is the RQ scale I’ve been working with:
40-49 Super Rare/C
50-64 Ultra Rare/B
In a perfect world, I would like a more consistent scale. I really think there’s an elegance and ease of understanding that comes with a new tier starting with every ten. I’m coming to accept there’s not a great way to do this though. There are around 25% more Ultra Rares than Super Rares, so it feels like that is the point to break the scale and jump to the 15s.
The new system means players start with 100 RQ instead of 35, and this is a little thing, but TD is often seen as daunting which repels a lot of potential players. It’s important we work to address this more. Starting at 100 RQ instead of 35 feels like one of those things that just seems a little less odd.
Currently we have a system where the RQ of a car within a rarity tier matters less the higher the tier, which theoretically means lower RQ cars in those tiers are less useful. The larger RQ bands for higher Rarities hopefully will increase the usefulness of those lower RQ cars within each rarity tier.
We don’t move the performance level of what it means to be an Epic or Legendary, e.g. the Lotus Exige S (2015) is often seen as the weakest Legendary. In general, when new cars come in, if they’re win less than that, they’re not Legendaries, if they’re stronger than that, they are. Cars can drift with engine changes or corrections, of course, but in general the barometer of what it means to be a particular rarity doesn’t change.
If we look at RQ values, the Lotus Exige S vs. the best Epic will be cheaper to use than it currently is, relatively speaking. 80/79 is 1.2% more RQ, currently it's 27/26, or 3.8% more RQ. So in the new system, you'll be getting the same performance boost for picking a Legendary, but the cost is less RQ.
One small change is when there were fewer ranks, we split up the cars relatively evenly within each rarity. This time, we’re using the winrate value itself to determine its RQ rank. This means some RQs have more cars than others, but it feels more authentic and appropriate for the game now. The difference of winrate between two RQ values values is smaller in those larger tiers - this is how there can be very few cars rated on each of the 95-100 RQ ranks.
In numbers this means two thirds of Legendary rarity cars are 80-89, versus one third 90-99. Even doing this, some Bugattis end up with the same RQ value. As one example, the current draft of RQ 98 has these cars (ordered randomly between the 5):
Bugatti Veyron 16.4 (2005)
McLaren P1 GTR (2015)
Bugatti Veyron 16.4 Super Sport (2010)
Lamborghini Veneno (2013)
Lamborghini Centenario (2016)
The Bugatti Veyron 16.4 Grand Sport (2009) sits at 97, the Bugatti Chiron (2017) is at 99, and the Bugatti Chiron Sport (2019) at 100, so pushing one car to another rank just creates another instance of this phenomenon at a different RQ level. This won’t happen nearly as often as it does now, but it’s not avoidable. (I suppose we could have even more ranks, but eventually the RQ difference becomes quite large between similar performing models).
We can now finely tune cars’ RQ more than before. In general, I don’t like to jump in and manually adjust cars’ RQ. We have changed some cars, but as a percentage, this is a very small number. The RQ system is always going to be bumpy - that’s a great thing. Players might think a car is under/overrated if the ratio of tracks in the RQ ranking tool ever deviates from what we use in events, or if scoring points is better for certain challenges vs. others, or if strategy means you use a car only when its strongest. We only need to tweak the cars that deviate so far in strength from that RQ level that it becomes disruptive.
Overall, the current RQ ranking system is flavour that’s very useful in the game. It means within each rarity tier, there’s more diversity than there would otherwise be, and it means there are hidden gems to be found. That is part of the strategy and fun of the game.
On that note, the top cars raised as needing the Specialist RQ tweaks (that we very occasionally do make) are the Chevrolet Colorado ZR2 and the Caterham Superlight R500. The suggested changes are to change their RQ and not their Rarity, as they have a long standing reputation, and so the fusing advantages would be very large (vs. the Land Rover changes where we assessed the advantage to be much smaller, and so did make the change to rarity). I think we’ll manually make them 64 and 79 RQ to start (the largest values within their rarity) and then we can think about any changes for future updates.
(By the way, I think the most likely outcome of a brand new RQ ranking system is balance of the game that is different...but not necessarily better or worse.) I’m not looking to fundamentally change the ranking system at all. The RQ changes should be just moving things to a new scale, rather than shaking the foundations of a system everyone is comfortable with and generally works well. This lowers the risk - the main concern is just hoping everyone can get used to the system within 2-3 days!