| ebay.es | n=16 | $2.32 | +126% |
| ebay.de | n=7 | $1.44 | +40% |
| ebay.co.uk | n=3 | $1.30 | +26% |
| ebay.com | n=22 | $1.03 | — |
How is this value calculated?
Recency-weighted trimmed median of 20 NM sales (RAW; outliers trimmed against the median, newer sales weigh more, dynamic half-life).
| Evidence path | Sales | Age | Value vᵢ | Weight gᵢ |
|---|---|---|---|---|
| eigene rohe NM-Verkäufe · Set -12% nachgeführt | 22 | 2 mo | $1.18 | 179.59 |
Measured error for this era×language (backtest): ±16 %
| Evidence path | Sales | Age | Value vᵢ | Weight gᵢ |
|---|---|---|---|---|
| en roh NM × 0.8701 (Sprach-Faktor aus 64 Karten) | 22 | 2 mo | $1.12 | 2.6 |
Measured error for this era×language (backtest): ±19 %
| Evidence path | Sales | Age | Value vᵢ | Weight gᵢ |
|---|---|---|---|---|
| en roh NM × 1.0231 (Sprach-Faktor aus 41 Karten) | 22 | 2 mo | $1.31 | 1.26 |
Measured error for this era×language (backtest): ±25 %
Gate not green — shown only here, never as the price.
| Evidence path | Sales | Age | Value vᵢ | Weight gᵢ |
|---|---|---|---|---|
| en roh NM × 1.0927 (Sprach-Faktor aus 30 Karten) · Set -15% nachgeführt | 22 | 2 mo | $1.27 | 1.95 |
Measured error for this era×language (backtest): ±17 %
No PSA numbers for this print yet — in any language.
#2 · 
#26 · 
#107 · 
#53 · 
#9
#91 · 
#63 · 
#42 · 
#35Frequently asked questions
What will I get for Dratini Obsidian Flames 157/197 if I want to sell quickly?
Half of all sales fell between $1.03 and $1.99, with a median of $1.48. Selling fast lands you nearer the lower end; with patience, good photos and a complete title the upper end is achievable.
Below which price is an offer for Dratini Obsidian Flames 157/197 unrealistic?
The cheapest genuine sales were around $0.97. Offers well below $0.58 are typically fakes, mislabelled cards or a different variant — ask for photos of all four borders and a close-up of the surface.
How does lastsold determine the market value?
lastsold uses only real sales — eBay plus auction houses like Goldin & Fanatics — recent sales count more (recency weighting), outliers and fakes are removed. It's not an estimate but actual sold prices, separated by condition, language and grading.