Preface
I wrote this preface when the rest of the article was already written. I must admit that this article was written somewhat out of my own annoyance. In the end, however, I did capture some data that might be helpful to players, or might spur further investigation (by more capable mathematicians :D).
Even though equal probabilities are usually something fair, I am very often blessed with bad luck in games that is second to none. I don't think I'm alone in this, and a luck factor is of course part of most games. In Splinterlands there are many variables that significantly influence the gameplay and that the player should take into account. Most of them are calculable and that gives even the luck in Splinterlands a tactical component. For example, the probability of missing an attack (which seems like luck in the first place) depends on a variety of abilities and stats of the individual cards. Both the speed of the monsters involved and a variety of abilities play a significant role. Among them are Flying, Blind, Dodge, Phase, etc.
If you want to find more information about this, you can read HERE about the individual probabilities of the abilities that affect missing, as well as HERE to see a sample calculation of how to calculate the value to miss.
So far so good, all fair. But there is another important luck component that you can't influence and that just annoys players with a lot of bad luck. The turn order in which the monsters attack. Why is it unfair to calculate this with luck? The outcome of a battle between two identical teams, which under normal circumstances should reach a draw, is decided solely by luck. Of course, the game has now found a solution for this simple circumstance: If two identical teams are actually facing each other, a draw is automatically triggered. This solves the principle at least in simple variants, but not in more complex games.
So if you were to create a match-up in which only two Sand Worm at the same level were facing each other, only luck would decide who starts first and wins accordingly, where by all logic it should be a draw if the cards are the same.
Experiment
Of course, I was bored enough to test it out a few times.
The first setup looked like this:
- Challenger mode - Novice League
- Alric Stormbringer vs Bortus (Because both don't change anything in the game, but make the game possible, because we get different teams through them)
- Gold foil Sand Worm (Alric) vs Sand Worm (Bortus) (both level 1. one is gold foil to cover another possible indicator)
- Always request the battle with Alric and be the first to accept the battle
- Only score who attacks first to override potential rules (Win = First Attack).
Result / Wins:
Bortus with Sand Worm: 7
Alric with Gold Sand Worm: 14
Ok I have only done a small test series and it is therefore of course not meaningful. The number of wins can be largely determined by luck. However, since I consider the difference in win rates of 66.6% (14 wins) to 33.3% (7 wins) to be very large, but I don't have time to test thousands of games, I decided to do a cross-check.
Possible parameters that can significantly influence the "luck" of the player in my experiment are A) Gold Foil brings more luck than Regular Foil, B) Alric has an advantage over Bortus, C) The speed in which a fight is accepted gives a luck advantage, D) The one who gives the Invite has an advantage.
All these parameters can be checked with the following changes in the setup, and then derive single experiments if necessary:
- Challenger mode - Novice League
- Alric Stormbringer vs Bortus
- Gold foil Sand Worm (Bortus) vs Sand Worm (Alric)
- Always request the battle with Bortus and accept the battle first
- Only score who attacks first to override potential rules
Result / Wins:
Bortus with Gold Sand Worm: 13
Alric with Sand Worm: 8
Alright, after this further pass we see fairly similar numbers. So there seems to be at least a chance that one of the factors mentioned has an influence on the player's luck.
In the following I will now check the individual parameters A-D. Parameter B can be deleted first of all, since we have disproved our assumption in the cross-check. Bortus won here more often than Alric.
The following setup can be used to check whether parameter A or parameter C+D have an influence on the player's luck:
- Challenger mode - Novice League
- Alric Stormbringer vs Bortus
- Gold foil Sand Worm (Alric) vs Sand Worm (Bortus)
- Always request the battle with Bortus and accept the battle first
- Only score who attacks first to override potential rules
Result / Wins:
Bortus with Sand Worm: 13
Alric with Gold Sand Worm: 8
We again have the same values as in the crosscheck, so we can assume about a 62%-38% Win ratio, rather than a 66%-33% Win ratio. We have found that the choice of cards as Gold or Regular Foil has no role on the draw order of identical cards.
The last test run is a very important one to be able to check the applicability in online PVP. This is because it is still possible that parameter D is decisive. Parameter D only refers to the Challenge mode (Invite a Friend to Fight. Not ranked), in which I tested the games.
For this the test setup looks like this:
- Challenger mode - Novice League
- Alric Stormbringer vs Bortus
- Gold foil Sand Worm (Alric) vs Sand Worm (Bortus)
- Always request the battle with Bortus, but accept the battle with Alric first
- Only score who attacks first, to override potential rules
Result / Wins:
Bortus with Sand Worm: 13
Alric with Gold Sand Worm: 8
Aaaaand dödödöm....
I really thought we were on to something big that could give us an advantage in online PvP. Unfortunately, my test only proved that the "host" of a game in challenge mode seems to have an advantage over the invited player.
Of course, much more data would have to be collected to get a clear result. A total of 84 games tested is not remotely reliable. However, since I have already put more than enough time into the test and there are no other parameters to test, I would end the experiment at this point and draw a summary.
Summary
The result of the test has shown that there is probably at least in the Challenger mode a parameter that influences "luck" in the turn order of the cards. It is obvious that such a parameter is also present in the online game, e.g. latency or similar, but it is not within the scope of my possibilities to conduct a test on this. It would require a very large amount of data and would ideally be tested with a bot and not by hand.
The luck factor in the turn order seems to be very difficult or even impossible for players to influence. What sounds good at first, however, can also be very frustrating, as described in the first paragraphs of the article, especially in an otherwise tactical game like Splinterlands. It would be desirable to either reveal the parameters for "luck" to the player, allowing them to adjust their tactics, or remove non-influenceable luck in the game to provide the necessary fairness.
What do the results do for us now?
Until someone has done more far-reaching tests and thus proven any parameters for luck in turn order, we can safely ignore it and dismiss it as annoying. This is also an advantage because you don't have to hope for unproven parameters that influence luck in the game. So feel free to take your time choosing your monsters and don't buy Gold Foils just because you think they have turn priority. By the way, it probably won't do you any good to press the shoulder buttons of your Gameboy just to make Mewtu finally stay in the ball... oops wrong game.
Attached you'll find a summary with all collected numbers in comparison, in case it helps anyone:
Result / Wins:
Bortus: 46
Alric: 38
Gold Sand Worm: 43
Sand Worm: 41
Invited to a battle: 53
First accept in battle: 48
If you are not yet a member of the Splinterlands community, feel free to join via my affiliate link. When you buy the Spell Book I will then get a part of the purchase price :)
https://splinterlands.com?ref=toheadi