We have all felt like every time we bring certain color heros, we get tiles that we don’t need and end up with bad boards. We think the Dev’s are “out to get us” with bad boards. Others have done great research on this topic.
I have ran through 750 mono team and 83 other team combos raids to gather the statistics see if this is the case. I recorded the starting boards of all 823 raids to see if the color tiles are skewed. All of these raids are done from v21 to present version (Oct 2019, yes finally got a chance to post the results). Due to the lack of heros, I was only able to run 3 mono teams (Blue, Purple and Green). Each mono team remained mostly the same throughout the 250 raids
With 35 tiles on the board and 5 colors, it is expected on average you will get 7 of each color tile. After 823 raids, and over 28K tiles, the tiles totals match the expect totals. Totals were within 0.05% - 0.32%
If I only take into account the 3 mono teams then the tiles recieved matches the expect tiles even more closely. Totals were within 0.01% - 0.35%
For the Blue Mono Team. I always had a feeling I was being shorted blue tiles for the whole 250 raids. For the first 50 raids, the stats did prove the case. After 100 raids, the numbers even out and matched expected occurences. It dip slightly by 150 raids but by 200 raids average blue tile occurence match expectancy again. I think this is what occuring with anybody that feels they are being shorted. We all remember the bad boards and not the good boards. This also illustrates, you can have a run of 50 bad raids but eventually it averages back out.
Of the expected 1750 blue tiles for 250 raids, I was only shorted 14 tiles, which is 0.16% or roughly losing 1 blue tie every 17.9 raids, so practically nothing.
Distribution of tiles looks normal. 96 times I had less than the average 7 blue ties and 97 times i had more than 7 tiles.
I attacked 117 Red Tank teams and average Blue tiles was 6.7 tiles per raid. Against a Green Tank (116 attacks) the Blue tiles average was 7.0 tiles. So there is little difference between the strong and weak color tanks for the Blue team
For the Purple Mono Team. I did not had the of being shorted, even though I ended up with a few less tiles than compare to my Blue Mono team.
Distribution of tiles looks to normal. 103 times I had less than the average 7 blue ties and 97 times i had more than 7 tiles.
I attacked yellow tank teams 66x and average tile was purple 7.0 tiles. For a purple tank team, I attack 188x and average 6.9 purple tiles. Just like the Blue mono team, there is little difference in Purple mono team against it’s weak strong and weak color tank teams.
For the Mono Green team, I had a feeling that I was getting good amount of green times, but this end up not the case. Of the 3 Mono Teams, my green team was the most shorted of its color. I think because I winning most of its raids, it gave me a sense of feeling that I was getting a lot of green tiles. But the truth was it was the strong team that was helping me win and not the strong tile number.
Of the expected 1750 blue tiles for 250 raids, I was only shorted 60 tiles, which is 0.69% or roughly losing 1 purple tie every 4.2 raids, so practically nothing
Distribution of tiles looks normal. 140 times I had less than the average 110 blue ties and 97 times i had more than 7 tiles which is relected in the shortage of green tiles.
In conclusion. There is no color bias for raids! It’s all in our heads 
My guess is war attacks will have the same results as raids. For the distribution stats may be someone with a strong math background can enlighten us if the distrubtion stats are within exceptable ranges
