`2024_19_GB_PHI` has different drive result with what on ESPN.com

[ak47twq]

test<-pbp%>%
  filter(game_id=="2024_19_GB_PHI") %>%
  select(play_id,play_type,desc,fixed_drive,fixed_drive_result) %>%
  collect()

while in ESPN play_by_play

https://www.espn.com/nfl/playbyplay/_/gameId/401671880

I think the 1st drive which result was FUMBLE is missing in PBP data.

Is it on purpose or a bug?

[guga31bb]

This is intentional (https://github.com/nflverse/nflfastR/blob/master/R/helper_add_fixed_drives.R#L17) but don't feel strongly here

To me it feels weird to call what the Packers had a drive if they never had the ball. If it were a recovered onsides kick would ESPN give them a drive?

[ak47twq]

they do not give a drive for recovered onside kick

https://www.espn.com/nfl/playbyplay/_/gameId/401547501

[ak47twq]

i see your point. but still feels weird to me that 2024_19_GB_PHI starts with drive no. 2.
when i do drive chart myself. the pic looks strange.

drive no. 1 is empty in the pic.

[mrcaseb]

I would like to keep the logic as is since it's not really different from an onside kick. However, I understand that starting with drive no. 2 is weird. I see 51 games since 2001 that start with drive no. 2 after a fumbled kickoff or recovered onside kick to start the game. Maybe we can handle this specific situation and decrement fixed drive @guga31bb ?

[guga31bb]

Fine with me!

[mrcaseb]

I think we could use this logic with row == 2

nflfastR/R/helper_add_fixed_drives.R

Lines 90 to 91 in 7d53e07

	# first observation of a half is also a new drive
	new_drive = dplyr::if_else(.data$row == 1, 1, .data$new_drive),

Something along

new_drive = dplyr::if_else(.data$row == 2, 0, .data$new_drive)

[mrcaseb]

I think we could use this logic with row == 2

nflfastR/R/helper_add_fixed_drives.R

Lines 90 to 91 in 7d53e07

first observation of a half is also a new drive

new_drive = dplyr::if_else(.data$row == 1, 1, .data$new_drive),
Something along

new_drive = dplyr::if_else(.data$row == 2, 0, .data$new_drive)

Actually if we do that, we end up with changing posteams and non changing fixed drive?