What may be modelled with the 10m error is not the CEP radius, but twice the radius. CEP only covers 50% of hits, twice the radius almost 95%. By this metric GLONASS guided weapons should have a 20m error value however.
But I’ll test in a bit to see the hit distribution and whether it can hit outside this error value (due to guidance of course) or not. I’m thinking it will be an uniform distribution, which is of course not correct…
I also wonder what affect range has.
I dropped 4x PW4s onto a destroyer from about 6km away, spreading them down the deck like I would in SB and 50% missed.
Id maybe beleive 10m of deviation for a long ranged drop or a lofted bomb, but seems like a lot of deviation even at relatively short distances.
Gaijin likes realism but only when it’s easy to model
So apparently the spread is actually rectangular (with sides equal to 2x the error value), not circular… tested on a completely flat map. Maus for reference (roughly 10m in length).
A-10C and GBU-39:
Groups with arrows were low altitude drops at “fast” and short ranges.
Other groups are dropped straight down.
F-15E and GBU-39:
Top group was launched from far and high, rightside.
Rest were low and fast from directions shown.
F-15E and GBU-38:
Right group was launched from far and high.
Rest were low and fast from directions shown.
Spread pattern seems affected by impact angle, so low and fast/short drops vs high altitude and far drops. It’s possible the error includes height which may explain it, otherwise it may simply just be the guidance not having enough time to adjust. It’s likely that the bomb tries to hit with a random coordinate inside a cube (with sides of 2x the error value) aligned with the world coordinates. Distribution seems to be uniform, but it’s quite weird at times (like in the A-10C top left group).
Thank you for doing these experiments, the distribution seems to be quite low-variance as well, and it looks like something is keeping bombs from landing closer to the target.
Yeah a lot of things in game are square. Like, fun fact: your missile’s seeker gimbal is a square, despite showing up as a circle. So you can actually get a bit more off-boresight performance on some missiles in specific situations by rotating 45 degrees.
It’s such a stupid change when GPS is probably the most widely known uncertainty of 5m.
Pretty much all this does is make it so small GNSS weapons like GBU-39s are unless against everything more armour than a truck.
I get that this was probably made so that the new SAMs aren’t made useless by GBU-39 spam, but they introduced the 2000lb JDAM/ERs in THE SAME UPDATE with a kill radius only slightly smaller than the inaccuracy.
If they want to keep this, they should at least reduce it to the realistic 5m so I can still use the GBU-39s in sim.
GBU-39s are now completely useless as the blast range is much less than the amount of deviation the game gives them. Pure luck now.
Out of curiosity, do the misses line up with the angle of approach and if yes do they always overshoot or under shoot?
If i remember correctly there was an issue previously with gps munitions that were guided via mission target selection and/or map point selection on some maps as the point was X meters in the air making the munitions always overshoot. Just a complete guess but it looks like that might be a possible cause here as well.
Either way, you can bug report it here (just search a bit first so that you don’t report something that is already reported and known about): (Gaijin.net // Issues)
A guide on how to report bugs can be found here: ([Navigation] Technical Knowledge Base | War Thunder Wiki)
Unguided weapons are incapable of countering current SPAA threats. It should also be noted that there are only a handful of guided munitions that cause problems in-game.
They both under and overshoot IIRC. Groups with an arrow are those with a low angle of approach in the direction of that arrow.
The SPI was just set in 3rd person, so maybe that issue with the map selection still exists seperately. I believe the target point is chosen in a cube now centered on the SPI with sides of 2x the error value, which is the “radius” of the cube (10m in this case giving a 20x20x20m cube). So they can target a point underground and thus undershoot as they collide with the ground trying to reach it, and overshoot when it’s targeting a point above the ground, inside the cube.
I am not sure if there’s anything to bug report, because it seems to work as intended. Just that the accuracy is not realistic.
Looks like GPS guided munitions have been tweaked as of 2.47.0.47, although it appears to be much wider in drop radius than in real life
aaaand the original bug report stating the spread should be AT MOST 3.3 meters has been marked as fixed, despite the spread being much larger at 5-7m
GBU -39 should have 1m or less too
https://community.gaijin.net/issues/p/warthunder/i/LiJtxtVaRvgQ
Good to see it was already accepted
Yeah, noticed it just now
I hope something actually happens with it.
A lot of reports are accepted…
Few are actually used.
I really want to send 10km pinpoint little paper plane bombs at clueless Pantsirs again.
I mean very few reports are simpler then this, all it takes is changing a single value in a few files, hopefully they get around to it sooner then later cause it shouldn’t be hard to fix once they do.
We do have the Data to change the 39, the Bug report got accepted, the change was promised under the post and not brought into the Game whatsoever, even though the Data is correct, backed up and the Report was accepted.
Average bug report activities.
True
