# Gravity strength in first person games

In our game, we set the acceleration due to gravity to the appropriate value of 9.8 m/s^2. Although physically correct, it feels more like moving around on the moon.. It feels about right at a value of 16 m/s^2. Has anyone else experienced this? Is it only because we don't have motion blur, or sound to give the apparent feeling of speed? Is it OK to just to set it to 16 m/s^2 if it feels right?

• Good question by the way, very common to see this, a lot of people assume they have to change the gravity when they encounter this – brandon May 4 '11 at 13:40
• Something that seems to have been assumed - are your object masses correct? – DrDeth May 4 '11 at 14:54
• mass shouldn't be taken into account unless you are simulating planets. A marble falls with the same acceleration as a bowling ball ignoring wind resistance – brandon May 4 '11 at 14:59
• Is your jump speed also realistic? If you've got 9.8m/s^2 gravity but a human who jumps 100x higher, the gravity will feel very wrong. Also, are you sure that you're applying the acceleration at the appropriate time scale? – Gregory Avery-Weir May 6 '11 at 16:42

It could be that your scale is off. If you are looking at a brick and it seems to fall slower than it should, you may just have it scaled larger than a brick should be. I had this problem when I was working the physX engine. My entire scene was about 3x too big so I scaled it down and it worked.

If 16 seems better than 9.8 but you want to use 9.8. Just multiply the size of the object falling by 0.61. Which should scale it down to where 9.8 gives you the desired effect.

Edit: To fix this. I made an object that should be 1 meter tall. I scale it until it seems to fall at the right speed. Whatever the world units are compared to the new scale gives me my new unit scale. Example. If the object was scaled to 0.14 then that represents 1 meter in the game world. Make a ruler object that is .14 tall and when you design the level you will be able to use that ruler to scale everything correctly. That will cause everything to look like it's falling correctly at 9.8 m/ss.

• That doesn't sound right. If you use consistent units, an object at 1 meter high should fall to the ground in 0.45 seconds. Scaling your entire scene shouldn't matter. Maybe you were communicating your timescale wrongly to your physics engine. – Kasper May 4 '11 at 14:16
• I'm not talking about the height that you drop it at. I'm talking about scaling the object itself. If you are a 100 foot man and you jump at 20% of your height. It will take you longer and feel like you are on the moon. The other side is, if you are a 6 foot man and jump at 20% of your height, you will jump much lower and will hit the ground much sooner. It is definitely a scale issue. – brandon May 4 '11 at 14:42

I'd say go with whatever feels and plays right. However, it would be better if you were using the correct value - if only so you or other developers don't have a WTF moment when you revisit the code 6 months down the line.

Don't forget that your physics implementation is only ever going to be an approximation of the real world and that sometimes getting things "too close" to reality makes the game unplayable. Also there may be something else in your simulation that's "off". Double check that as well.

More than a few years ago I was involved with a helicopter simulator and the client wanted the controls to be "realistic". It failed because it was unplayable. Making the controls simpler made the game playable.

Modelling everything in the metric system I also observed that gravity feels much too weak with regards to the player jumping (in a platformer). Much like "on the moon", "jelly", "too much phyics", not "arcade" enough. On the other hand players in games are regularly able to jump to great unrealistic height too.

TLDR: high gravity feels very fast paced and arcade, realistic gravity sometimes might not give the desired game experience.