Testing out some different strategies using Archmage Bot
Hi everyone! I recently came back to Splinterlands after an 8-month hiatus, and one of the biggest changes in the third-party space is bots that are easy to use and pretty high quality, notable Archmage. Soon after hearing about it, I picked one up and have been pretty happy with it. In this time of meta shakeup, I definitely play better than it, but it plays when I'm asleep. And burnout is what contributed to my hiatus - I couldn't keep playing every day and keeping my ECR low. So I mostly moved everything into SPS and picked up the passive profits from that. Archmage will obviate that problem and use my ECR if I go on vacation, need a break from the game, etc.
One question I've been toying with is how to maximize the bot. It can play anywhere from 1-4 times per hour. At once an hour, because you use 1% ECR but regenerate 1.04%, you will miss 1 full battle's worth of ECR per day. At twice an hour, it will trend down towards 50% but never go below that. At three times an hour, it will go down to 50% after 2 days. And at 4 times an hour it will hit 50% in just 30 hours.
Per the Archmage devs, 2x/hr with a limit anywhere between 1 and 0.5 will work in perpetuity, as shown in the image above. This will maximize Rewards. But, I don't think it will maximize Rating, and this is something I want to test.
Rating, unlike Rewards does not depend on ECR and just on your rating and your opponent's rating. Because of this, more battles is strictly better than fewer battles for a given win%, as you will occasionally get win streaks boosting your rating, and that will more often (in an absolute, not a relative sense) with more battles.
So, my goal is to test out a few different strategies to see which balances chest rewards and ranking by the end of the season. Each of the next 2 seasons I'll be trying out a different strategy and updating with the results.
Strategy 1: Staying close to 98% ECR the whole season.
The first strategy will be the one I think will be worst: leaving the bot set at 2x/hr with a minimum ECR of 0.98. This will result in ECR remaining at ~0.99 the whole season for a total of ~25*14=350 battles during the season.
Strategy 2: Staying at 50% the whole season.
The second strategy will be more aggressive. It will start at 100% ECR and do 2x battles per hour for the entire season. Because it will not quite go down to 0.5, I might do a couple additional battles at the very end of the season. In total, this method should result in 672 battles during the season (it isn't 2x the previous version because 25% ECR is regenerated every day, so this method could be a little more aggressive in playing down to exactly 50%).
Those two strategies should illustrate my main point here: despite both maximizing ECR Rewards, I think only the latter will maximize Ranking Points. And I think the resulting number and expected value of chests will be higher for the latter strategy.
Conclusions and Next Steps
In short, I hope to show that more battles per hour, while equal for rewards, is better for ranking. Importantly, while I'll be showing this with a bot, the same holds true if you're manually piloting an account. The bot merely allows it be a little more regular and controllable.
After these strategies, I might try some more complicated approaches to eke out a little more. These might include ignoring the first two days of the season while the top ranked people fight it out. This would forsake the beginning of season rewards which are usually higher, but it would avoid spending significant ECR while not moving up rapidly. I don't know whether this will be better or not.
Interesting in Splinterlands? Interested in testing out the archmage bot? Use my referral link! https://splinterlands.com?ref=badrag