Race predictor is quite a bit off...

Mine is off but in the opposite direction, it predicts faster than I could possibly run...some distances it predicts significantly quicker times than distance PB's I obtained 4 years ago, when I was training much more consistently and focussed on being able to run quicker.

I get this across all predictors...Stryd is even further off especially in the longer distances (even when I give it actual courses to work with) - possibly because I don't do formal CP tests or run maximal efforts very often in training, RunAlyze after being quite far off (too quick OR too slow) for the last couple of years actually at the moment seems to be showing me a reasonable prognosis of what my best forecast distance efforts might possibly be now (based on recent 10k race and best effort parkrun times I've achieved where I've felt I've ran close to 'as quickly as I reasonably can' at the mo) - it feels like it's VERY SLIGHTLY slow across all distances compared to what I think I can achieve - the odd minute at 10k, 10m, Half and maybe a couple of mins slow on Marathon prognosis (hopefully - its predicting 4h2m and I'd really like to (and feel I can, if everything goes well) go sub 4hrs again for the first time in ages next weekend!!).

My approach is to use a bit of gut instinct, refined from B races and other hard efforts, and knowing what a hard parkrun time is...knock a bit of time off the Runalyze prognosis, and build a Garmin PacePro plan (with updates on elevation changes, so I don't get too hung up every km when I'm failing!) to hit that.

Apparently, your watch under estimate your pace at VO2Max. The best thing you can do is to have a balanced training, and an accurate HRMax value. Make sure your races and intense workouts are taking your stamina to the floor.

widert said:

I've been running quite a lot of races lately, but the race predictor is quite a bit off:

How is your VO2Max trend and HRV status? If strained, that could explain why your VO2Max is under estimated these days.

All models come with margin of errors, and applied results depend on the quality of data coming in. Garmin's essentials parameters are your VO2Max, your age, sex and weight, and your training history (which influences how well estimated your VO2Max is).

Note 1:

Garmin predictions work great for me, slightly optimistic, but realistic

Note 2:

Your HM predictions are worse than your 5k predictions. I plugged your data in 2 of the many race predictors you can find online. I used the (a) Jeff Galloway method, because it works great for me. I also used this (b) calculator (I like it because it gives you training paces as well)

Race Equivalency Calculator (lukehumphreyrunning.com)

Based on your 5k time, (a) predicts 1:27:03 and (b) predict 1:27:01 for the HM (97.7% of your actual speed). For a marathon, model (a) predicts 3:08:36, and model (b) predicts 3:01:26.

Therefore, Garmin estimates your 5k speed at 96.8% of your actual, and your HM speed at 91.4% of your actual